{ "scope": "Synthetic diagnostic, not held-out production evaluation. Hosted Jev not tested.", "experiment_date": "2026-09-22", "source_revision": "90990a5fac2995b9faa3190f7d437e84f2067768", "kev_configuration": "Kev-4B / Tesla T4 / FP16 / unmerged adapter; causal_conv1d fallback", "limitations": [ "Overlapping categories", "Different semantic instructions", "Rules developed alongside examples", "Deferral threshold selected post hoc" ], "gold_label_matches": { "rules": 58, "gpt54mini": 59, "kev": 55, "gliclass": 13 }, "kev_latency_ms": { "first": 84225.91, "subsequent_median": 147.1, "subsequent_mean": 209.04745762711863 }, "posthoc_deferral": { "margin_at_least": 0.5, "accepted": 41, "accepted_correct": 41, "deferred": 19 }, "decisions": [ { "model_id": "rules", "case_id": "social-attribution-v1-000", "prompt": "I'm long NHX here and adding on any dip under forty. This is my highest-conviction name for the quarter.", "expected_label": "explicit_position", "predicted_label": "explicit_position", "latency_ms": 0.05, "scores": null, "correct": true }, { "model_id": "rules", "case_id": "social-attribution-v1-001", "prompt": "Closed my entire copper position this morning. I'd rather sit in cash than fight this tape.", "expected_label": "explicit_position", "predicted_label": "unclear", "latency_ms": 0.05, "scores": null, "correct": false }, { "model_id": "rules", "case_id": "social-attribution-v1-002", "prompt": "My take: the regional banks are uninvestable right now and I won't touch them.", "expected_label": "explicit_position", "predicted_label": "explicit_position", "latency_ms": 0.04, "scores": null, "correct": true }, { "model_id": "rules", "case_id": "social-attribution-v1-003", "prompt": "I bought the dip in the semis basket today. Full size, no hedge, and I'm comfortable holding through earnings.", "expected_label": "explicit_position", "predicted_label": "explicit_position", "latency_ms": 0.04, "scores": null, "correct": true }, { "model_id": "rules", "case_id": "social-attribution-v1-004", "prompt": "Staying overweight energy. I've said it for weeks and nothing in this print changes my mind.", "expected_label": "explicit_position", "predicted_label": "explicit_position", "latency_ms": 0.04, "scores": null, "correct": true }, { "model_id": "rules", "case_id": "social-attribution-v1-005", "prompt": "I trimmed half of my gold miners into the rip. Locking gains, still hold a starter.", "expected_label": "explicit_position", "predicted_label": "explicit_position", "latency_ms": 0.04, "scores": null, "correct": true }, { "model_id": "rules", "case_id": "social-attribution-v1-006", "prompt": "Short the homebuilders. That's where my book is positioned and I'm sizing it up into the meeting.", "expected_label": "explicit_position", "predicted_label": "explicit_position", "latency_ms": 0.04, "scores": null, "correct": true }, { "model_id": "rules", "case_id": "social-attribution-v1-007", "prompt": "I'm flat everything into the number. No positions, no exposure, watching from the sideline by choice.", "expected_label": "explicit_position", "predicted_label": "explicit_position", "latency_ms": 0.04, "scores": null, "correct": true }, { "model_id": "rules", "case_id": "social-attribution-v1-008", "prompt": "Been accumulating the uranium names all month and I'll keep buying weakness. Conviction is high.", "expected_label": "explicit_position", "predicted_label": "explicit_position", "latency_ms": 0.04, "scores": null, "correct": true }, { "model_id": "rules", "case_id": "social-attribution-v1-009", "prompt": "I sold my airlines yesterday and I'm not looking back. Wrong thesis, cut it, moving on.", "expected_label": "explicit_position", "predicted_label": "explicit_position", "latency_ms": 0.04, "scores": null, "correct": true }, { "model_id": "rules", "case_id": "social-attribution-v1-010", "prompt": "If VLR reclaims the 200-day on volume, I'll start a position. Until then I do nothing.", "expected_label": "conditional", "predicted_label": "conditional", "latency_ms": 0.04, "scores": null, "correct": true }, { "model_id": "rules", "case_id": "social-attribution-v1-011", "prompt": "Should the jobs number come in soft, I'd add to my rate-cut trades. It hasn't yet, so no move.", "expected_label": "conditional", "predicted_label": "conditional", "latency_ms": 0.05, "scores": null, "correct": true }, { "model_id": "rules", "case_id": "social-attribution-v1-012", "prompt": "As long as crude holds above seventy, the tanker thesis stays alive for me. Lose it and I'm out.", "expected_label": "conditional", "predicted_label": "conditional", "latency_ms": 0.04, "scores": null, "correct": true }, { "model_id": "rules", "case_id": "social-attribution-v1-013", "prompt": "I'll get bullish the moment this downtrend line breaks. Not before, not on hope.", "expected_label": "conditional", "predicted_label": "conditional", "latency_ms": 0.04, "scores": null, "correct": true }, { "model_id": "rules", "case_id": "social-attribution-v1-014", "prompt": "When the VIX finally cracks under fifteen I'll redeploy the cash pile. We're not there.", "expected_label": "conditional", "predicted_label": "conditional", "latency_ms": 0.04, "scores": null, "correct": true }, { "model_id": "rules", "case_id": "social-attribution-v1-015", "prompt": "Provided the guidance isn't cut, I stay in. A cut and I reassess the whole position overnight.", "expected_label": "conditional", "predicted_label": "conditional", "latency_ms": 0.04, "scores": null, "correct": true }, { "model_id": "rules", "case_id": "social-attribution-v1-016", "prompt": "If they lose that support shelf I flip short in a heartbeat. Watching it tick by tick.", "expected_label": "conditional", "predicted_label": "conditional", "latency_ms": 0.04, "scores": null, "correct": true }, { "model_id": "rules", "case_id": "social-attribution-v1-017", "prompt": "Only if the spread tightens back under two hundred would I touch the high-yield names again.", "expected_label": "conditional", "predicted_label": "conditional", "latency_ms": 0.04, "scores": null, "correct": true }, { "model_id": "rules", "case_id": "social-attribution-v1-018", "prompt": "Assuming the merger clears review, the arb is worth it. If it breaks, this trade is dead money.", "expected_label": "conditional", "predicted_label": "conditional", "latency_ms": 0.04, "scores": null, "correct": true }, { "model_id": "rules", "case_id": "social-attribution-v1-019", "prompt": "I'd chase the breakout, but only on a daily close above the prior high. Intraday pokes don't count.", "expected_label": "conditional", "predicted_label": "conditional", "latency_ms": 0.04, "scores": null, "correct": true }, { "model_id": "rules", "case_id": "social-attribution-v1-020", "prompt": "The central bank left rates unchanged and signaled one more hike is possible before year-end.", "expected_label": "news", "predicted_label": "news", "latency_ms": 0.03, "scores": null, "correct": true }, { "model_id": "rules", "case_id": "social-attribution-v1-021", "prompt": "Quarterly revenue came in at 4.2 billion, up nine percent from a year ago, the company reported.", "expected_label": "news", "predicted_label": "news", "latency_ms": 0.04, "scores": null, "correct": true }, { "model_id": "rules", "case_id": "social-attribution-v1-022", "prompt": "Regulators approved the acquisition this morning, clearing the last major hurdle for the deal.", "expected_label": "news", "predicted_label": "news", "latency_ms": 0.03, "scores": null, "correct": true }, { "model_id": "rules", "case_id": "social-attribution-v1-023", "prompt": "Oil settled two percent lower after inventories rose more than analysts had expected.", "expected_label": "news", "predicted_label": "news", "latency_ms": 0.03, "scores": null, "correct": true }, { "model_id": "rules", "case_id": "social-attribution-v1-024", "prompt": "The index closed at a record high, its fourth straight gain, on broad participation across sectors.", "expected_label": "news", "predicted_label": "news", "latency_ms": 0.03, "scores": null, "correct": true }, { "model_id": "rules", "case_id": "social-attribution-v1-025", "prompt": "The company announced a recall of two of its product lines pending a safety review.", "expected_label": "news", "predicted_label": "news", "latency_ms": 0.03, "scores": null, "correct": true }, { "model_id": "rules", "case_id": "social-attribution-v1-026", "prompt": "Payrolls grew by 180,000 last month and the unemployment rate ticked up to 4.1 percent.", "expected_label": "news", "predicted_label": "news", "latency_ms": 0.03, "scores": null, "correct": true }, { "model_id": "rules", "case_id": "social-attribution-v1-027", "prompt": "A major ratings agency downgraded the sovereign one notch, citing rising debt-service costs.", "expected_label": "news", "predicted_label": "news", "latency_ms": 0.03, "scores": null, "correct": true }, { "model_id": "rules", "case_id": "social-attribution-v1-028", "prompt": "The chipmaker said it will build a new fabrication plant, its largest capital outlay to date.", "expected_label": "news", "predicted_label": "unclear", "latency_ms": 0.05, "scores": null, "correct": false }, { "model_id": "rules", "case_id": "social-attribution-v1-029", "prompt": "Freight volumes fell for a third consecutive month, according to the industry group's latest data.", "expected_label": "news", "predicted_label": "news", "latency_ms": 0.03, "scores": null, "correct": true }, { "model_id": "rules", "case_id": "social-attribution-v1-030", "prompt": "A viewer asks whether I think lithium has bottomed. Good question, and honestly I'm not the one to call it.", "expected_label": "quoted", "predicted_label": "quoted", "latency_ms": 0.02, "scores": null, "correct": true }, { "model_id": "rules", "case_id": "social-attribution-v1-031", "prompt": "The strategist on the panel argued small caps outperform from here. That's her view, not a claim of mine.", "expected_label": "quoted", "predicted_label": "quoted", "latency_ms": 0.02, "scores": null, "correct": true }, { "model_id": "rules", "case_id": "social-attribution-v1-032", "prompt": "Someone in the chat swears this name doubles by spring. I'm just passing along what they said.", "expected_label": "quoted", "predicted_label": "quoted", "latency_ms": 0.02, "scores": null, "correct": true }, { "model_id": "rules", "case_id": "social-attribution-v1-033", "prompt": "My mentor always told me to fade the first gap. Sharing his rule, do with it what you will.", "expected_label": "quoted", "predicted_label": "quoted", "latency_ms": 0.03, "scores": null, "correct": true }, { "model_id": "rules", "case_id": "social-attribution-v1-034", "prompt": "A subscriber emailed that they're loading up on the utilities. Relaying it for the room to weigh.", "expected_label": "quoted", "predicted_label": "quoted", "latency_ms": 0.02, "scores": null, "correct": true }, { "model_id": "rules", "case_id": "social-attribution-v1-035", "prompt": "The famous investor said on stage he'd never short a dull market. Quoting him, not endorsing it.", "expected_label": "quoted", "predicted_label": "quoted", "latency_ms": 0.02, "scores": null, "correct": true }, { "model_id": "rules", "case_id": "social-attribution-v1-036", "prompt": "One of you asked if the dividend is safe. I put the question to the group rather than answer blind.", "expected_label": "quoted", "predicted_label": "quoted", "latency_ms": 0.02, "scores": null, "correct": true }, { "model_id": "rules", "case_id": "social-attribution-v1-037", "prompt": "A caller claimed the sector's already priced the recession. That's his framing; I'm reporting it.", "expected_label": "quoted", "predicted_label": "quoted", "latency_ms": 0.02, "scores": null, "correct": true }, { "model_id": "rules", "case_id": "social-attribution-v1-038", "prompt": "The analyst note making the rounds says fair value is ninety. Circulating it, form your own view.", "expected_label": "quoted", "predicted_label": "quoted", "latency_ms": 0.02, "scores": null, "correct": true }, { "model_id": "rules", "case_id": "social-attribution-v1-039", "prompt": "A follower DM'd that insiders are buying. I can't verify it, I'm only surfacing the tip.", "expected_label": "quoted", "predicted_label": "quoted", "latency_ms": 0.02, "scores": null, "correct": true }, { "model_id": "rules", "case_id": "social-attribution-v1-040", "prompt": "Doors to the premium alerts room close Friday. Lock in the founder rate before it goes up.", "expected_label": "promotion", "predicted_label": "promotion", "latency_ms": 0.01, "scores": null, "correct": true }, { "model_id": "rules", "case_id": "social-attribution-v1-041", "prompt": "Grab my free options-flow checklist at the link. Drops straight to your inbox, no strings.", "expected_label": "promotion", "predicted_label": "promotion", "latency_ms": 0.02, "scores": null, "correct": true }, { "model_id": "rules", "case_id": "social-attribution-v1-042", "prompt": "This breakdown is brought to you by my charting platform. Use code EDGE for two months free.", "expected_label": "promotion", "predicted_label": "promotion", "latency_ms": 0.01, "scores": null, "correct": true }, { "model_id": "rules", "case_id": "social-attribution-v1-043", "prompt": "Join the weekend workshop where I walk the full playbook live. Seats are limited, link below.", "expected_label": "promotion", "predicted_label": "promotion", "latency_ms": 0.01, "scores": null, "correct": true }, { "model_id": "rules", "case_id": "social-attribution-v1-044", "prompt": "My new course on position sizing is live today. Early birds get the bonus module at no cost.", "expected_label": "promotion", "predicted_label": "promotion", "latency_ms": 0.01, "scores": null, "correct": true }, { "model_id": "rules", "case_id": "social-attribution-v1-045", "prompt": "Subscribe to the pro tier for the full model and the daily watchlist. Cancel anytime, obviously.", "expected_label": "promotion", "predicted_label": "promotion", "latency_ms": 0.01, "scores": null, "correct": true }, { "model_id": "rules", "case_id": "social-attribution-v1-046", "prompt": "Sponsored: open an account with the broker in my bio and get commission-free trades for a year.", "expected_label": "promotion", "predicted_label": "promotion", "latency_ms": 0.01, "scores": null, "correct": true }, { "model_id": "rules", "case_id": "social-attribution-v1-047", "prompt": "Want the spreadsheet I use? It's in the members' vault. Membership link is pinned up top.", "expected_label": "promotion", "predicted_label": "promotion", "latency_ms": 0.01, "scores": null, "correct": true }, { "model_id": "rules", "case_id": "social-attribution-v1-048", "prompt": "Last call on the mentorship cohort. Apply tonight and I'll review your book one-on-one.", "expected_label": "promotion", "predicted_label": "promotion", "latency_ms": 0.01, "scores": null, "correct": true }, { "model_id": "rules", "case_id": "social-attribution-v1-049", "prompt": "Download the app and follow my live portfolio in real time. First month is on the house.", "expected_label": "promotion", "predicted_label": "promotion", "latency_ms": 0.01, "scores": null, "correct": true }, { "model_id": "rules", "case_id": "social-attribution-v1-050", "prompt": "Interesting setup on this one. We'll see how it plays out.", "expected_label": "unclear", "predicted_label": "unclear", "latency_ms": 0.04, "scores": null, "correct": true }, { "model_id": "rules", "case_id": "social-attribution-v1-051", "prompt": "Lots of chatter about the sector today. Hard to say which way it breaks from here.", "expected_label": "unclear", "predicted_label": "unclear", "latency_ms": 0.07, "scores": null, "correct": true }, { "model_id": "rules", "case_id": "social-attribution-v1-052", "prompt": "Well, that was a day. On to tomorrow.", "expected_label": "unclear", "predicted_label": "unclear", "latency_ms": 0.05, "scores": null, "correct": true }, { "model_id": "rules", "case_id": "social-attribution-v1-053", "prompt": "Something feels off about this whole move, but I can't put my finger on it yet.", "expected_label": "unclear", "predicted_label": "unclear", "latency_ms": 0.05, "scores": null, "correct": true }, { "model_id": "rules", "case_id": "social-attribution-v1-054", "prompt": "Charts, charts everywhere. Anyway, hope everyone had a decent session.", "expected_label": "unclear", "predicted_label": "unclear", "latency_ms": 0.03, "scores": null, "correct": true }, { "model_id": "rules", "case_id": "social-attribution-v1-055", "prompt": "Not sure what to make of the print. Could go either way, genuinely.", "expected_label": "unclear", "predicted_label": "unclear", "latency_ms": 0.04, "scores": null, "correct": true }, { "model_id": "rules", "case_id": "social-attribution-v1-056", "prompt": "A lot of moving parts here. Let it settle before deciding anything.", "expected_label": "unclear", "predicted_label": "unclear", "latency_ms": 0.04, "scores": null, "correct": true }, { "model_id": "rules", "case_id": "social-attribution-v1-057", "prompt": "Mixed signals across the board today. No strong read on it.", "expected_label": "unclear", "predicted_label": "unclear", "latency_ms": 0.04, "scores": null, "correct": true }, { "model_id": "rules", "case_id": "social-attribution-v1-058", "prompt": "That headline is either huge or nothing. Time will tell, I suppose.", "expected_label": "unclear", "predicted_label": "unclear", "latency_ms": 0.04, "scores": null, "correct": true }, { "model_id": "rules", "case_id": "social-attribution-v1-059", "prompt": "Watching, thinking, sipping coffee. The market does what it wants.", "expected_label": "unclear", "predicted_label": "unclear", "latency_ms": 0.04, "scores": null, "correct": true }, { "model_id": "gpt54mini", "case_id": "social-attribution-v1-000", "prompt": "I'm long NHX here and adding on any dip under forty. This is my highest-conviction name for the quarter.", "expected_label": "explicit_position", "predicted_label": "explicit_position", "latency_ms": 2640.54, "scores": null, "correct": true }, { "model_id": "gpt54mini", "case_id": "social-attribution-v1-001", "prompt": "Closed my entire copper position this morning. I'd rather sit in cash than fight this tape.", "expected_label": "explicit_position", "predicted_label": "explicit_position", "latency_ms": 1809.14, "scores": null, "correct": true }, { "model_id": "gpt54mini", "case_id": "social-attribution-v1-002", "prompt": "My take: the regional banks are uninvestable right now and I won't touch them.", "expected_label": "explicit_position", "predicted_label": "explicit_position", "latency_ms": 1832.16, "scores": null, "correct": true }, { "model_id": "gpt54mini", "case_id": "social-attribution-v1-003", "prompt": "I bought the dip in the semis basket today. Full size, no hedge, and I'm comfortable holding through earnings.", "expected_label": "explicit_position", "predicted_label": "explicit_position", "latency_ms": 2382.16, "scores": null, "correct": true }, { "model_id": "gpt54mini", "case_id": "social-attribution-v1-004", "prompt": "Staying overweight energy. I've said it for weeks and nothing in this print changes my mind.", "expected_label": "explicit_position", "predicted_label": "explicit_position", "latency_ms": 2050.57, "scores": null, "correct": true }, { "model_id": "gpt54mini", "case_id": "social-attribution-v1-005", "prompt": "I trimmed half of my gold miners into the rip. Locking gains, still hold a starter.", "expected_label": "explicit_position", "predicted_label": "explicit_position", "latency_ms": 1782.59, "scores": null, "correct": true }, { "model_id": "gpt54mini", "case_id": "social-attribution-v1-006", "prompt": "Short the homebuilders. That's where my book is positioned and I'm sizing it up into the meeting.", "expected_label": "explicit_position", "predicted_label": "explicit_position", "latency_ms": 2632.66, "scores": null, "correct": true }, { "model_id": "gpt54mini", "case_id": "social-attribution-v1-007", "prompt": "I'm flat everything into the number. No positions, no exposure, watching from the sideline by choice.", "expected_label": "explicit_position", "predicted_label": "explicit_position", "latency_ms": 2168.7, "scores": null, "correct": true }, { "model_id": "gpt54mini", "case_id": "social-attribution-v1-008", "prompt": "Been accumulating the uranium names all month and I'll keep buying weakness. Conviction is high.", "expected_label": "explicit_position", "predicted_label": "explicit_position", "latency_ms": 1804.96, "scores": null, "correct": true }, { "model_id": "gpt54mini", "case_id": "social-attribution-v1-009", "prompt": "I sold my airlines yesterday and I'm not looking back. Wrong thesis, cut it, moving on.", "expected_label": "explicit_position", "predicted_label": "explicit_position", "latency_ms": 2417.6, "scores": null, "correct": true }, { "model_id": "gpt54mini", "case_id": "social-attribution-v1-010", "prompt": "If VLR reclaims the 200-day on volume, I'll start a position. Until then I do nothing.", "expected_label": "conditional", "predicted_label": "conditional", "latency_ms": 2087.9, "scores": null, "correct": true }, { "model_id": "gpt54mini", "case_id": "social-attribution-v1-011", "prompt": "Should the jobs number come in soft, I'd add to my rate-cut trades. It hasn't yet, so no move.", "expected_label": "conditional", "predicted_label": "conditional", "latency_ms": 1627.54, "scores": null, "correct": true }, { "model_id": "gpt54mini", "case_id": "social-attribution-v1-012", "prompt": "As long as crude holds above seventy, the tanker thesis stays alive for me. Lose it and I'm out.", "expected_label": "conditional", "predicted_label": "conditional", "latency_ms": 2666.42, "scores": null, "correct": true }, { "model_id": "gpt54mini", "case_id": "social-attribution-v1-013", "prompt": "I'll get bullish the moment this downtrend line breaks. Not before, not on hope.", "expected_label": "conditional", "predicted_label": "conditional", "latency_ms": 2656.18, "scores": null, "correct": true }, { "model_id": "gpt54mini", "case_id": "social-attribution-v1-014", "prompt": "When the VIX finally cracks under fifteen I'll redeploy the cash pile. We're not there.", "expected_label": "conditional", "predicted_label": "conditional", "latency_ms": 1622.86, "scores": null, "correct": true }, { "model_id": "gpt54mini", "case_id": "social-attribution-v1-015", "prompt": "Provided the guidance isn't cut, I stay in. A cut and I reassess the whole position overnight.", "expected_label": "conditional", "predicted_label": "conditional", "latency_ms": 2372.48, "scores": null, "correct": true }, { "model_id": "gpt54mini", "case_id": "social-attribution-v1-016", "prompt": "If they lose that support shelf I flip short in a heartbeat. Watching it tick by tick.", "expected_label": "conditional", "predicted_label": "conditional", "latency_ms": 1463.15, "scores": null, "correct": true }, { "model_id": "gpt54mini", "case_id": "social-attribution-v1-017", "prompt": "Only if the spread tightens back under two hundred would I touch the high-yield names again.", "expected_label": "conditional", "predicted_label": "conditional", "latency_ms": 2040.48, "scores": null, "correct": true }, { "model_id": "gpt54mini", "case_id": "social-attribution-v1-018", "prompt": "Assuming the merger clears review, the arb is worth it. If it breaks, this trade is dead money.", "expected_label": "conditional", "predicted_label": "conditional", "latency_ms": 2263.28, "scores": null, "correct": true }, { "model_id": "gpt54mini", "case_id": "social-attribution-v1-019", "prompt": "I'd chase the breakout, but only on a daily close above the prior high. Intraday pokes don't count.", "expected_label": "conditional", "predicted_label": "conditional", "latency_ms": 2112.5, "scores": null, "correct": true }, { "model_id": "gpt54mini", "case_id": "social-attribution-v1-020", "prompt": "The central bank left rates unchanged and signaled one more hike is possible before year-end.", "expected_label": "news", "predicted_label": "news", "latency_ms": 1548.66, "scores": null, "correct": true }, { "model_id": "gpt54mini", "case_id": "social-attribution-v1-021", "prompt": "Quarterly revenue came in at 4.2 billion, up nine percent from a year ago, the company reported.", "expected_label": "news", "predicted_label": "news", "latency_ms": 2017.81, "scores": null, "correct": true }, { "model_id": "gpt54mini", "case_id": "social-attribution-v1-022", "prompt": "Regulators approved the acquisition this morning, clearing the last major hurdle for the deal.", "expected_label": "news", "predicted_label": "news", "latency_ms": 1943.88, "scores": null, "correct": true }, { "model_id": "gpt54mini", "case_id": "social-attribution-v1-023", "prompt": "Oil settled two percent lower after inventories rose more than analysts had expected.", "expected_label": "news", "predicted_label": "news", "latency_ms": 1467.61, "scores": null, "correct": true }, { "model_id": "gpt54mini", "case_id": "social-attribution-v1-024", "prompt": "The index closed at a record high, its fourth straight gain, on broad participation across sectors.", "expected_label": "news", "predicted_label": "news", "latency_ms": 1573.92, "scores": null, "correct": true }, { "model_id": "gpt54mini", "case_id": "social-attribution-v1-025", "prompt": "The company announced a recall of two of its product lines pending a safety review.", "expected_label": "news", "predicted_label": "news", "latency_ms": 2522.09, "scores": null, "correct": true }, { "model_id": "gpt54mini", "case_id": "social-attribution-v1-026", "prompt": "Payrolls grew by 180,000 last month and the unemployment rate ticked up to 4.1 percent.", "expected_label": "news", "predicted_label": "news", "latency_ms": 2508.92, "scores": null, "correct": true }, { "model_id": "gpt54mini", "case_id": "social-attribution-v1-027", "prompt": "A major ratings agency downgraded the sovereign one notch, citing rising debt-service costs.", "expected_label": "news", "predicted_label": "news", "latency_ms": 2399.03, "scores": null, "correct": true }, { "model_id": "gpt54mini", "case_id": "social-attribution-v1-028", "prompt": "The chipmaker said it will build a new fabrication plant, its largest capital outlay to date.", "expected_label": "news", "predicted_label": "news", "latency_ms": 2326.02, "scores": null, "correct": true }, { "model_id": "gpt54mini", "case_id": "social-attribution-v1-029", "prompt": "Freight volumes fell for a third consecutive month, according to the industry group's latest data.", "expected_label": "news", "predicted_label": "news", "latency_ms": 1410.41, "scores": null, "correct": true }, { "model_id": "gpt54mini", "case_id": "social-attribution-v1-030", "prompt": "A viewer asks whether I think lithium has bottomed. Good question, and honestly I'm not the one to call it.", "expected_label": "quoted", "predicted_label": "quoted", "latency_ms": 2857.55, "scores": null, "correct": true }, { "model_id": "gpt54mini", "case_id": "social-attribution-v1-031", "prompt": "The strategist on the panel argued small caps outperform from here. That's her view, not a claim of mine.", "expected_label": "quoted", "predicted_label": "quoted", "latency_ms": 2146.44, "scores": null, "correct": true }, { "model_id": "gpt54mini", "case_id": "social-attribution-v1-032", "prompt": "Someone in the chat swears this name doubles by spring. I'm just passing along what they said.", "expected_label": "quoted", "predicted_label": "quoted", "latency_ms": 2313.7, "scores": null, "correct": true }, { "model_id": "gpt54mini", "case_id": "social-attribution-v1-033", "prompt": "My mentor always told me to fade the first gap. Sharing his rule, do with it what you will.", "expected_label": "quoted", "predicted_label": "quoted", "latency_ms": 2202.56, "scores": null, "correct": true }, { "model_id": "gpt54mini", "case_id": "social-attribution-v1-034", "prompt": "A subscriber emailed that they're loading up on the utilities. Relaying it for the room to weigh.", "expected_label": "quoted", "predicted_label": "quoted", "latency_ms": 1959.01, "scores": null, "correct": true }, { "model_id": "gpt54mini", "case_id": "social-attribution-v1-035", "prompt": "The famous investor said on stage he'd never short a dull market. Quoting him, not endorsing it.", "expected_label": "quoted", "predicted_label": "quoted", "latency_ms": 3277.51, "scores": null, "correct": true }, { "model_id": "gpt54mini", "case_id": "social-attribution-v1-036", "prompt": "One of you asked if the dividend is safe. I put the question to the group rather than answer blind.", "expected_label": "quoted", "predicted_label": "quoted", "latency_ms": 2294.66, "scores": null, "correct": true }, { "model_id": "gpt54mini", "case_id": "social-attribution-v1-037", "prompt": "A caller claimed the sector's already priced the recession. That's his framing; I'm reporting it.", "expected_label": "quoted", "predicted_label": "quoted", "latency_ms": 2404.02, "scores": null, "correct": true }, { "model_id": "gpt54mini", "case_id": "social-attribution-v1-038", "prompt": "The analyst note making the rounds says fair value is ninety. Circulating it, form your own view.", "expected_label": "quoted", "predicted_label": "news", "latency_ms": 2798.35, "scores": null, "correct": false }, { "model_id": "gpt54mini", "case_id": "social-attribution-v1-039", "prompt": "A follower DM'd that insiders are buying. I can't verify it, I'm only surfacing the tip.", "expected_label": "quoted", "predicted_label": "quoted", "latency_ms": 2582.62, "scores": null, "correct": true }, { "model_id": "gpt54mini", "case_id": "social-attribution-v1-040", "prompt": "Doors to the premium alerts room close Friday. Lock in the founder rate before it goes up.", "expected_label": "promotion", "predicted_label": "promotion", "latency_ms": 2627.67, "scores": null, "correct": true }, { "model_id": "gpt54mini", "case_id": "social-attribution-v1-041", "prompt": "Grab my free options-flow checklist at the link. Drops straight to your inbox, no strings.", "expected_label": "promotion", "predicted_label": "promotion", "latency_ms": 1607.77, "scores": null, "correct": true }, { "model_id": "gpt54mini", "case_id": "social-attribution-v1-042", "prompt": "This breakdown is brought to you by my charting platform. Use code EDGE for two months free.", "expected_label": "promotion", "predicted_label": "promotion", "latency_ms": 2166.42, "scores": null, "correct": true }, { "model_id": "gpt54mini", "case_id": "social-attribution-v1-043", "prompt": "Join the weekend workshop where I walk the full playbook live. Seats are limited, link below.", "expected_label": "promotion", "predicted_label": "promotion", "latency_ms": 2722.6, "scores": null, "correct": true }, { "model_id": "gpt54mini", "case_id": "social-attribution-v1-044", "prompt": "My new course on position sizing is live today. Early birds get the bonus module at no cost.", "expected_label": "promotion", "predicted_label": "promotion", "latency_ms": 1603.73, "scores": null, "correct": true }, { "model_id": "gpt54mini", "case_id": "social-attribution-v1-045", "prompt": "Subscribe to the pro tier for the full model and the daily watchlist. Cancel anytime, obviously.", "expected_label": "promotion", "predicted_label": "promotion", "latency_ms": 2794.44, "scores": null, "correct": true }, { "model_id": "gpt54mini", "case_id": "social-attribution-v1-046", "prompt": "Sponsored: open an account with the broker in my bio and get commission-free trades for a year.", "expected_label": "promotion", "predicted_label": "promotion", "latency_ms": 1986.17, "scores": null, "correct": true }, { "model_id": "gpt54mini", "case_id": "social-attribution-v1-047", "prompt": "Want the spreadsheet I use? It's in the members' vault. Membership link is pinned up top.", "expected_label": "promotion", "predicted_label": "promotion", "latency_ms": 2191.98, "scores": null, "correct": true }, { "model_id": "gpt54mini", "case_id": "social-attribution-v1-048", "prompt": "Last call on the mentorship cohort. Apply tonight and I'll review your book one-on-one.", "expected_label": "promotion", "predicted_label": "promotion", "latency_ms": 2752.76, "scores": null, "correct": true }, { "model_id": "gpt54mini", "case_id": "social-attribution-v1-049", "prompt": "Download the app and follow my live portfolio in real time. First month is on the house.", "expected_label": "promotion", "predicted_label": "promotion", "latency_ms": 2109.36, "scores": null, "correct": true }, { "model_id": "gpt54mini", "case_id": "social-attribution-v1-050", "prompt": "Interesting setup on this one. We'll see how it plays out.", "expected_label": "unclear", "predicted_label": "unclear", "latency_ms": 2332.1, "scores": null, "correct": true }, { "model_id": "gpt54mini", "case_id": "social-attribution-v1-051", "prompt": "Lots of chatter about the sector today. Hard to say which way it breaks from here.", "expected_label": "unclear", "predicted_label": "unclear", "latency_ms": 2117.75, "scores": null, "correct": true }, { "model_id": "gpt54mini", "case_id": "social-attribution-v1-052", "prompt": "Well, that was a day. On to tomorrow.", "expected_label": "unclear", "predicted_label": "unclear", "latency_ms": 1725.87, "scores": null, "correct": true }, { "model_id": "gpt54mini", "case_id": "social-attribution-v1-053", "prompt": "Something feels off about this whole move, but I can't put my finger on it yet.", "expected_label": "unclear", "predicted_label": "unclear", "latency_ms": 2379.98, "scores": null, "correct": true }, { "model_id": "gpt54mini", "case_id": "social-attribution-v1-054", "prompt": "Charts, charts everywhere. Anyway, hope everyone had a decent session.", "expected_label": "unclear", "predicted_label": "unclear", "latency_ms": 2474.91, "scores": null, "correct": true }, { "model_id": "gpt54mini", "case_id": "social-attribution-v1-055", "prompt": "Not sure what to make of the print. Could go either way, genuinely.", "expected_label": "unclear", "predicted_label": "unclear", "latency_ms": 1659.27, "scores": null, "correct": true }, { "model_id": "gpt54mini", "case_id": "social-attribution-v1-056", "prompt": "A lot of moving parts here. Let it settle before deciding anything.", "expected_label": "unclear", "predicted_label": "unclear", "latency_ms": 1717.68, "scores": null, "correct": true }, { "model_id": "gpt54mini", "case_id": "social-attribution-v1-057", "prompt": "Mixed signals across the board today. No strong read on it.", "expected_label": "unclear", "predicted_label": "unclear", "latency_ms": 1735.55, "scores": null, "correct": true }, { "model_id": "gpt54mini", "case_id": "social-attribution-v1-058", "prompt": "That headline is either huge or nothing. Time will tell, I suppose.", "expected_label": "unclear", "predicted_label": "unclear", "latency_ms": 2630.07, "scores": null, "correct": true }, { "model_id": "gpt54mini", "case_id": "social-attribution-v1-059", "prompt": "Watching, thinking, sipping coffee. The market does what it wants.", "expected_label": "unclear", "predicted_label": "unclear", "latency_ms": 2498.08, "scores": null, "correct": true }, { "model_id": "kev", "case_id": "social-attribution-v1-000", "prompt": "I'm long NHX here and adding on any dip under forty. This is my highest-conviction name for the quarter.", "expected_label": "explicit_position", "predicted_label": "explicit_position", "latency_ms": 84225.91, "scores": { "conditional": 0.0486, "explicit_position": 0.8893, "news": 0.0126, "promotion": 0.0092, "quoted": 0.0162, "unclear": 0.024 }, "correct": true }, { "model_id": "kev", "case_id": "social-attribution-v1-001", "prompt": "Closed my entire copper position this morning. I'd rather sit in cash than fight this tape.", "expected_label": "explicit_position", "predicted_label": "explicit_position", "latency_ms": 405.46, "scores": { "conditional": 0.0641, "explicit_position": 0.554, "news": 0.0882, "promotion": 0.0277, "quoted": 0.1284, "unclear": 0.1376 }, "correct": true }, { "model_id": "kev", "case_id": "social-attribution-v1-002", "prompt": "My take: the regional banks are uninvestable right now and I won't touch them.", "expected_label": "explicit_position", "predicted_label": "explicit_position", "latency_ms": 146.17, "scores": { "conditional": 0.0176, "explicit_position": 0.931, "news": 0.0129, "promotion": 0.0049, "quoted": 0.0175, "unclear": 0.0161 }, "correct": true }, { "model_id": "kev", "case_id": "social-attribution-v1-003", "prompt": "I bought the dip in the semis basket today. Full size, no hedge, and I'm comfortable holding through earnings.", "expected_label": "explicit_position", "predicted_label": "explicit_position", "latency_ms": 390.82, "scores": { "conditional": 0.0347, "explicit_position": 0.8602, "news": 0.0352, "promotion": 0.0216, "quoted": 0.0193, "unclear": 0.029 }, "correct": true }, { "model_id": "kev", "case_id": "social-attribution-v1-004", "prompt": "Staying overweight energy. I've said it for weeks and nothing in this print changes my mind.", "expected_label": "explicit_position", "predicted_label": "explicit_position", "latency_ms": 390.96, "scores": { "conditional": 0.0813, "explicit_position": 0.5078, "news": 0.0671, "promotion": 0.0707, "quoted": 0.0825, "unclear": 0.1905 }, "correct": true }, { "model_id": "kev", "case_id": "social-attribution-v1-005", "prompt": "I trimmed half of my gold miners into the rip. Locking gains, still hold a starter.", "expected_label": "explicit_position", "predicted_label": "explicit_position", "latency_ms": 145.89, "scores": { "conditional": 0.0226, "explicit_position": 0.8933, "news": 0.0169, "promotion": 0.008, "quoted": 0.0204, "unclear": 0.0387 }, "correct": true }, { "model_id": "kev", "case_id": "social-attribution-v1-006", "prompt": "Short the homebuilders. That's where my book is positioned and I'm sizing it up into the meeting.", "expected_label": "explicit_position", "predicted_label": "explicit_position", "latency_ms": 390.68, "scores": { "conditional": 0.0512, "explicit_position": 0.7118, "news": 0.0519, "promotion": 0.0402, "quoted": 0.0464, "unclear": 0.0986 }, "correct": true }, { "model_id": "kev", "case_id": "social-attribution-v1-007", "prompt": "I'm flat everything into the number. No positions, no exposure, watching from the sideline by choice.", "expected_label": "explicit_position", "predicted_label": "explicit_position", "latency_ms": 390.71, "scores": { "conditional": 0.092, "explicit_position": 0.3496, "news": 0.1491, "promotion": 0.0329, "quoted": 0.1138, "unclear": 0.2625 }, "correct": true }, { "model_id": "kev", "case_id": "social-attribution-v1-008", "prompt": "Been accumulating the uranium names all month and I'll keep buying weakness. Conviction is high.", "expected_label": "explicit_position", "predicted_label": "explicit_position", "latency_ms": 145.52, "scores": { "conditional": 0.0252, "explicit_position": 0.8898, "news": 0.0169, "promotion": 0.0106, "quoted": 0.0243, "unclear": 0.0332 }, "correct": true }, { "model_id": "kev", "case_id": "social-attribution-v1-009", "prompt": "I sold my airlines yesterday and I'm not looking back. Wrong thesis, cut it, moving on.", "expected_label": "explicit_position", "predicted_label": "explicit_position", "latency_ms": 146.29, "scores": { "conditional": 0.0256, "explicit_position": 0.7734, "news": 0.0549, "promotion": 0.0261, "quoted": 0.0499, "unclear": 0.0701 }, "correct": true }, { "model_id": "kev", "case_id": "social-attribution-v1-010", "prompt": "If VLR reclaims the 200-day on volume, I'll start a position. Until then I do nothing.", "expected_label": "conditional", "predicted_label": "conditional", "latency_ms": 147.7, "scores": { "conditional": 0.9356, "explicit_position": 0.0295, "news": 0.0065, "promotion": 0.0025, "quoted": 0.0124, "unclear": 0.0135 }, "correct": true }, { "model_id": "kev", "case_id": "social-attribution-v1-011", "prompt": "Should the jobs number come in soft, I'd add to my rate-cut trades. It hasn't yet, so no move.", "expected_label": "conditional", "predicted_label": "explicit_position", "latency_ms": 148.02, "scores": { "conditional": 0.3779, "explicit_position": 0.5899, "news": 0.0089, "promotion": 0.0023, "quoted": 0.0076, "unclear": 0.0134 }, "correct": false }, { "model_id": "kev", "case_id": "social-attribution-v1-012", "prompt": "As long as crude holds above seventy, the tanker thesis stays alive for me. Lose it and I'm out.", "expected_label": "conditional", "predicted_label": "conditional", "latency_ms": 391.28, "scores": { "conditional": 0.8267, "explicit_position": 0.1249, "news": 0.0099, "promotion": 0.0047, "quoted": 0.0152, "unclear": 0.0185 }, "correct": true }, { "model_id": "kev", "case_id": "social-attribution-v1-013", "prompt": "I'll get bullish the moment this downtrend line breaks. Not before, not on hope.", "expected_label": "conditional", "predicted_label": "conditional", "latency_ms": 145.53, "scores": { "conditional": 0.9222, "explicit_position": 0.0394, "news": 0.0053, "promotion": 0.0036, "quoted": 0.0149, "unclear": 0.0144 }, "correct": true }, { "model_id": "kev", "case_id": "social-attribution-v1-014", "prompt": "When the VIX finally cracks under fifteen I'll redeploy the cash pile. We're not there.", "expected_label": "conditional", "predicted_label": "conditional", "latency_ms": 143.81, "scores": { "conditional": 0.9423, "explicit_position": 0.0268, "news": 0.0069, "promotion": 0.0029, "quoted": 0.0082, "unclear": 0.0129 }, "correct": true }, { "model_id": "kev", "case_id": "social-attribution-v1-015", "prompt": "Provided the guidance isn't cut, I stay in. A cut and I reassess the whole position overnight.", "expected_label": "conditional", "predicted_label": "conditional", "latency_ms": 146.63, "scores": { "conditional": 0.6853, "explicit_position": 0.1398, "news": 0.0264, "promotion": 0.0179, "quoted": 0.0639, "unclear": 0.0668 }, "correct": true }, { "model_id": "kev", "case_id": "social-attribution-v1-016", "prompt": "If they lose that support shelf I flip short in a heartbeat. Watching it tick by tick.", "expected_label": "conditional", "predicted_label": "conditional", "latency_ms": 145.77, "scores": { "conditional": 0.6053, "explicit_position": 0.1693, "news": 0.037, "promotion": 0.0149, "quoted": 0.0896, "unclear": 0.0839 }, "correct": true }, { "model_id": "kev", "case_id": "social-attribution-v1-017", "prompt": "Only if the spread tightens back under two hundred would I touch the high-yield names again.", "expected_label": "conditional", "predicted_label": "conditional", "latency_ms": 146.73, "scores": { "conditional": 0.9625, "explicit_position": 0.0127, "news": 0.0045, "promotion": 0.0023, "quoted": 0.0087, "unclear": 0.0094 }, "correct": true }, { "model_id": "kev", "case_id": "social-attribution-v1-018", "prompt": "Assuming the merger clears review, the arb is worth it. If it breaks, this trade is dead money.", "expected_label": "conditional", "predicted_label": "conditional", "latency_ms": 146.25, "scores": { "conditional": 0.8586, "explicit_position": 0.0556, "news": 0.0205, "promotion": 0.009, "quoted": 0.0184, "unclear": 0.0378 }, "correct": true }, { "model_id": "kev", "case_id": "social-attribution-v1-019", "prompt": "I'd chase the breakout, but only on a daily close above the prior high. Intraday pokes don't count.", "expected_label": "conditional", "predicted_label": "conditional", "latency_ms": 148.36, "scores": { "conditional": 0.6786, "explicit_position": 0.1696, "news": 0.0211, "promotion": 0.0119, "quoted": 0.0535, "unclear": 0.0653 }, "correct": true }, { "model_id": "kev", "case_id": "social-attribution-v1-020", "prompt": "The central bank left rates unchanged and signaled one more hike is possible before year-end.", "expected_label": "news", "predicted_label": "news", "latency_ms": 392.24, "scores": { "conditional": 0.0202, "explicit_position": 0.0129, "news": 0.9225, "promotion": 0.0041, "quoted": 0.0174, "unclear": 0.0229 }, "correct": true }, { "model_id": "kev", "case_id": "social-attribution-v1-021", "prompt": "Quarterly revenue came in at 4.2 billion, up nine percent from a year ago, the company reported.", "expected_label": "news", "predicted_label": "news", "latency_ms": 391.57, "scores": { "conditional": 0.009, "explicit_position": 0.0146, "news": 0.892, "promotion": 0.0136, "quoted": 0.0246, "unclear": 0.0463 }, "correct": true }, { "model_id": "kev", "case_id": "social-attribution-v1-022", "prompt": "Regulators approved the acquisition this morning, clearing the last major hurdle for the deal.", "expected_label": "news", "predicted_label": "news", "latency_ms": 145.5, "scores": { "conditional": 0.0047, "explicit_position": 0.0056, "news": 0.9702, "promotion": 0.0035, "quoted": 0.0065, "unclear": 0.0096 }, "correct": true }, { "model_id": "kev", "case_id": "social-attribution-v1-023", "prompt": "Oil settled two percent lower after inventories rose more than analysts had expected.", "expected_label": "news", "predicted_label": "news", "latency_ms": 389.63, "scores": { "conditional": 0.0055, "explicit_position": 0.013, "news": 0.9411, "promotion": 0.0047, "quoted": 0.01, "unclear": 0.0258 }, "correct": true }, { "model_id": "kev", "case_id": "social-attribution-v1-024", "prompt": "The index closed at a record high, its fourth straight gain, on broad participation across sectors.", "expected_label": "news", "predicted_label": "news", "latency_ms": 146.24, "scores": { "conditional": 0.0039, "explicit_position": 0.0082, "news": 0.9594, "promotion": 0.0048, "quoted": 0.0072, "unclear": 0.0165 }, "correct": true }, { "model_id": "kev", "case_id": "social-attribution-v1-025", "prompt": "The company announced a recall of two of its product lines pending a safety review.", "expected_label": "news", "predicted_label": "news", "latency_ms": 389.56, "scores": { "conditional": 0.0051, "explicit_position": 0.0055, "news": 0.9632, "promotion": 0.0036, "quoted": 0.008, "unclear": 0.0146 }, "correct": true }, { "model_id": "kev", "case_id": "social-attribution-v1-026", "prompt": "Payrolls grew by 180,000 last month and the unemployment rate ticked up to 4.1 percent.", "expected_label": "news", "predicted_label": "news", "latency_ms": 394.23, "scores": { "conditional": 0.0042, "explicit_position": 0.0069, "news": 0.9518, "promotion": 0.0048, "quoted": 0.0101, "unclear": 0.0221 }, "correct": true }, { "model_id": "kev", "case_id": "social-attribution-v1-027", "prompt": "A major ratings agency downgraded the sovereign one notch, citing rising debt-service costs.", "expected_label": "news", "predicted_label": "news", "latency_ms": 146.18, "scores": { "conditional": 0.0065, "explicit_position": 0.0121, "news": 0.9268, "promotion": 0.0052, "quoted": 0.0182, "unclear": 0.0313 }, "correct": true }, { "model_id": "kev", "case_id": "social-attribution-v1-028", "prompt": "The chipmaker said it will build a new fabrication plant, its largest capital outlay to date.", "expected_label": "news", "predicted_label": "news", "latency_ms": 146.23, "scores": { "conditional": 0.0085, "explicit_position": 0.0092, "news": 0.9309, "promotion": 0.008, "quoted": 0.0198, "unclear": 0.0236 }, "correct": true }, { "model_id": "kev", "case_id": "social-attribution-v1-029", "prompt": "Freight volumes fell for a third consecutive month, according to the industry group's latest data.", "expected_label": "news", "predicted_label": "news", "latency_ms": 146.35, "scores": { "conditional": 0.0075, "explicit_position": 0.0165, "news": 0.9185, "promotion": 0.0058, "quoted": 0.0237, "unclear": 0.028 }, "correct": true }, { "model_id": "kev", "case_id": "social-attribution-v1-030", "prompt": "A viewer asks whether I think lithium has bottomed. Good question, and honestly I'm not the one to call it.", "expected_label": "quoted", "predicted_label": "quoted", "latency_ms": 147.41, "scores": { "conditional": 0.0745, "explicit_position": 0.113, "news": 0.0634, "promotion": 0.0219, "quoted": 0.4288, "unclear": 0.2984 }, "correct": true }, { "model_id": "kev", "case_id": "social-attribution-v1-031", "prompt": "The strategist on the panel argued small caps outperform from here. That's her view, not a claim of mine.", "expected_label": "quoted", "predicted_label": "quoted", "latency_ms": 147.66, "scores": { "conditional": 0.0031, "explicit_position": 0.0123, "news": 0.0155, "promotion": 0.0022, "quoted": 0.9565, "unclear": 0.0104 }, "correct": true }, { "model_id": "kev", "case_id": "social-attribution-v1-032", "prompt": "Someone in the chat swears this name doubles by spring. I'm just passing along what they said.", "expected_label": "quoted", "predicted_label": "quoted", "latency_ms": 146.58, "scores": { "conditional": 0.0209, "explicit_position": 0.0041, "news": 0.0104, "promotion": 0.0021, "quoted": 0.9489, "unclear": 0.0137 }, "correct": true }, { "model_id": "kev", "case_id": "social-attribution-v1-033", "prompt": "My mentor always told me to fade the first gap. Sharing his rule, do with it what you will.", "expected_label": "quoted", "predicted_label": "quoted", "latency_ms": 146.87, "scores": { "conditional": 0.0174, "explicit_position": 0.0289, "news": 0.0166, "promotion": 0.0071, "quoted": 0.8634, "unclear": 0.0667 }, "correct": true }, { "model_id": "kev", "case_id": "social-attribution-v1-034", "prompt": "A subscriber emailed that they're loading up on the utilities. Relaying it for the room to weigh.", "expected_label": "quoted", "predicted_label": "quoted", "latency_ms": 147.49, "scores": { "conditional": 0.0096, "explicit_position": 0.0107, "news": 0.1359, "promotion": 0.0075, "quoted": 0.7995, "unclear": 0.0368 }, "correct": true }, { "model_id": "kev", "case_id": "social-attribution-v1-035", "prompt": "The famous investor said on stage he'd never short a dull market. Quoting him, not endorsing it.", "expected_label": "quoted", "predicted_label": "quoted", "latency_ms": 148.8, "scores": { "conditional": 0.006, "explicit_position": 0.009, "news": 0.2162, "promotion": 0.0042, "quoted": 0.7348, "unclear": 0.0299 }, "correct": true }, { "model_id": "kev", "case_id": "social-attribution-v1-036", "prompt": "One of you asked if the dividend is safe. I put the question to the group rather than answer blind.", "expected_label": "quoted", "predicted_label": "quoted", "latency_ms": 147.55, "scores": { "conditional": 0.098, "explicit_position": 0.1242, "news": 0.1185, "promotion": 0.0432, "quoted": 0.434, "unclear": 0.1822 }, "correct": true }, { "model_id": "kev", "case_id": "social-attribution-v1-037", "prompt": "A caller claimed the sector's already priced the recession. That's his framing; I'm reporting it.", "expected_label": "quoted", "predicted_label": "quoted", "latency_ms": 147.52, "scores": { "conditional": 0.0039, "explicit_position": 0.0081, "news": 0.1625, "promotion": 0.0026, "quoted": 0.8028, "unclear": 0.0202 }, "correct": true }, { "model_id": "kev", "case_id": "social-attribution-v1-038", "prompt": "The analyst note making the rounds says fair value is ninety. Circulating it, form your own view.", "expected_label": "quoted", "predicted_label": "quoted", "latency_ms": 147.98, "scores": { "conditional": 0.0436, "explicit_position": 0.0919, "news": 0.2919, "promotion": 0.0402, "quoted": 0.3021, "unclear": 0.2304 }, "correct": true }, { "model_id": "kev", "case_id": "social-attribution-v1-039", "prompt": "A follower DM'd that insiders are buying. I can't verify it, I'm only surfacing the tip.", "expected_label": "quoted", "predicted_label": "quoted", "latency_ms": 147.43, "scores": { "conditional": 0.0256, "explicit_position": 0.0245, "news": 0.1154, "promotion": 0.0173, "quoted": 0.6395, "unclear": 0.1777 }, "correct": true }, { "model_id": "kev", "case_id": "social-attribution-v1-040", "prompt": "Doors to the premium alerts room close Friday. Lock in the founder rate before it goes up.", "expected_label": "promotion", "predicted_label": "news", "latency_ms": 148.29, "scores": { "conditional": 0.0269, "explicit_position": 0.0153, "news": 0.5931, "promotion": 0.321, "quoted": 0.0127, "unclear": 0.031 }, "correct": false }, { "model_id": "kev", "case_id": "social-attribution-v1-041", "prompt": "Grab my free options-flow checklist at the link. Drops straight to your inbox, no strings.", "expected_label": "promotion", "predicted_label": "promotion", "latency_ms": 147.01, "scores": { "conditional": 0.0521, "explicit_position": 0.0294, "news": 0.0311, "promotion": 0.7143, "quoted": 0.0202, "unclear": 0.1529 }, "correct": true }, { "model_id": "kev", "case_id": "social-attribution-v1-042", "prompt": "This breakdown is brought to you by my charting platform. Use code EDGE for two months free.", "expected_label": "promotion", "predicted_label": "promotion", "latency_ms": 147.1, "scores": { "conditional": 0.0048, "explicit_position": 0.006, "news": 0.0089, "promotion": 0.964, "quoted": 0.0047, "unclear": 0.0116 }, "correct": true }, { "model_id": "kev", "case_id": "social-attribution-v1-043", "prompt": "Join the weekend workshop where I walk the full playbook live. Seats are limited, link below.", "expected_label": "promotion", "predicted_label": "promotion", "latency_ms": 146.78, "scores": { "conditional": 0.0367, "explicit_position": 0.0179, "news": 0.0653, "promotion": 0.7158, "quoted": 0.0205, "unclear": 0.1438 }, "correct": true }, { "model_id": "kev", "case_id": "social-attribution-v1-044", "prompt": "My new course on position sizing is live today. Early birds get the bonus module at no cost.", "expected_label": "promotion", "predicted_label": "promotion", "latency_ms": 145.59, "scores": { "conditional": 0.0175, "explicit_position": 0.0062, "news": 0.0291, "promotion": 0.9291, "quoted": 0.0032, "unclear": 0.015 }, "correct": true }, { "model_id": "kev", "case_id": "social-attribution-v1-045", "prompt": "Subscribe to the pro tier for the full model and the daily watchlist. Cancel anytime, obviously.", "expected_label": "promotion", "predicted_label": "promotion", "latency_ms": 146.43, "scores": { "conditional": 0.0715, "explicit_position": 0.0125, "news": 0.0218, "promotion": 0.8513, "quoted": 0.0139, "unclear": 0.029 }, "correct": true }, { "model_id": "kev", "case_id": "social-attribution-v1-046", "prompt": "Sponsored: open an account with the broker in my bio and get commission-free trades for a year.", "expected_label": "promotion", "predicted_label": "promotion", "latency_ms": 147.84, "scores": { "conditional": 0.0071, "explicit_position": 0.0039, "news": 0.0046, "promotion": 0.972, "quoted": 0.0035, "unclear": 0.0089 }, "correct": true }, { "model_id": "kev", "case_id": "social-attribution-v1-047", "prompt": "Want the spreadsheet I use? It's in the members' vault. Membership link is pinned up top.", "expected_label": "promotion", "predicted_label": "unclear", "latency_ms": 146.69, "scores": { "conditional": 0.1457, "explicit_position": 0.0969, "news": 0.132, "promotion": 0.2493, "quoted": 0.0827, "unclear": 0.2935 }, "correct": false }, { "model_id": "kev", "case_id": "social-attribution-v1-048", "prompt": "Last call on the mentorship cohort. Apply tonight and I'll review your book one-on-one.", "expected_label": "promotion", "predicted_label": "promotion", "latency_ms": 147.91, "scores": { "conditional": 0.1674, "explicit_position": 0.0314, "news": 0.0718, "promotion": 0.4594, "quoted": 0.0408, "unclear": 0.2292 }, "correct": true }, { "model_id": "kev", "case_id": "social-attribution-v1-049", "prompt": "Download the app and follow my live portfolio in real time. First month is on the house.", "expected_label": "promotion", "predicted_label": "promotion", "latency_ms": 146.18, "scores": { "conditional": 0.0634, "explicit_position": 0.0541, "news": 0.0229, "promotion": 0.7846, "quoted": 0.0164, "unclear": 0.0586 }, "correct": true }, { "model_id": "kev", "case_id": "social-attribution-v1-050", "prompt": "Interesting setup on this one. We'll see how it plays out.", "expected_label": "unclear", "predicted_label": "unclear", "latency_ms": 390.83, "scores": { "conditional": 0.3308, "explicit_position": 0.1024, "news": 0.1259, "promotion": 0.0259, "quoted": 0.061, "unclear": 0.3541 }, "correct": true }, { "model_id": "kev", "case_id": "social-attribution-v1-051", "prompt": "Lots of chatter about the sector today. Hard to say which way it breaks from here.", "expected_label": "unclear", "predicted_label": "unclear", "latency_ms": 391.16, "scores": { "conditional": 0.0321, "explicit_position": 0.042, "news": 0.1152, "promotion": 0.0124, "quoted": 0.0351, "unclear": 0.7632 }, "correct": true }, { "model_id": "kev", "case_id": "social-attribution-v1-052", "prompt": "Well, that was a day. On to tomorrow.", "expected_label": "unclear", "predicted_label": "unclear", "latency_ms": 390.01, "scores": { "conditional": 0.0702, "explicit_position": 0.1017, "news": 0.2922, "promotion": 0.0452, "quoted": 0.1021, "unclear": 0.3886 }, "correct": true }, { "model_id": "kev", "case_id": "social-attribution-v1-053", "prompt": "Something feels off about this whole move, but I can't put my finger on it yet.", "expected_label": "unclear", "predicted_label": "unclear", "latency_ms": 146.33, "scores": { "conditional": 0.0869, "explicit_position": 0.1596, "news": 0.1003, "promotion": 0.0354, "quoted": 0.0734, "unclear": 0.5444 }, "correct": true }, { "model_id": "kev", "case_id": "social-attribution-v1-054", "prompt": "Charts, charts everywhere. Anyway, hope everyone had a decent session.", "expected_label": "unclear", "predicted_label": "unclear", "latency_ms": 146.66, "scores": { "conditional": 0.0585, "explicit_position": 0.0961, "news": 0.2165, "promotion": 0.0603, "quoted": 0.1017, "unclear": 0.467 }, "correct": true }, { "model_id": "kev", "case_id": "social-attribution-v1-055", "prompt": "Not sure what to make of the print. Could go either way, genuinely.", "expected_label": "unclear", "predicted_label": "unclear", "latency_ms": 146.42, "scores": { "conditional": 0.0149, "explicit_position": 0.0323, "news": 0.0119, "promotion": 0.009, "quoted": 0.0167, "unclear": 0.9152 }, "correct": true }, { "model_id": "kev", "case_id": "social-attribution-v1-056", "prompt": "A lot of moving parts here. Let it settle before deciding anything.", "expected_label": "unclear", "predicted_label": "conditional", "latency_ms": 145.13, "scores": { "conditional": 0.5674, "explicit_position": 0.1011, "news": 0.0491, "promotion": 0.0134, "quoted": 0.0331, "unclear": 0.2359 }, "correct": false }, { "model_id": "kev", "case_id": "social-attribution-v1-057", "prompt": "Mixed signals across the board today. No strong read on it.", "expected_label": "unclear", "predicted_label": "unclear", "latency_ms": 390.83, "scores": { "conditional": 0.0114, "explicit_position": 0.0221, "news": 0.0612, "promotion": 0.0069, "quoted": 0.0182, "unclear": 0.8803 }, "correct": true }, { "model_id": "kev", "case_id": "social-attribution-v1-058", "prompt": "That headline is either huge or nothing. Time will tell, I suppose.", "expected_label": "unclear", "predicted_label": "conditional", "latency_ms": 145.59, "scores": { "conditional": 0.4011, "explicit_position": 0.083, "news": 0.1034, "promotion": 0.014, "quoted": 0.0409, "unclear": 0.3577 }, "correct": false }, { "model_id": "kev", "case_id": "social-attribution-v1-059", "prompt": "Watching, thinking, sipping coffee. The market does what it wants.", "expected_label": "unclear", "predicted_label": "unclear", "latency_ms": 145.42, "scores": { "conditional": 0.0603, "explicit_position": 0.1477, "news": 0.2477, "promotion": 0.0413, "quoted": 0.1467, "unclear": 0.3562 }, "correct": true }, { "model_id": "gliclass", "case_id": "social-attribution-v1-000", "prompt": "I'm long NHX here and adding on any dip under forty. This is my highest-conviction name for the quarter.", "expected_label": "explicit_position", "predicted_label": "quoted", "latency_ms": 252.45, "scores": { "conditional": 0.03624030202627182, "explicit_position": 0.08689630031585693, "news": 0.017445355653762817, "promotion": 0.01749519631266594, "quoted": 0.8335925936698914, "unclear": 0.008330346085131168 }, "correct": false }, { "model_id": "gliclass", "case_id": "social-attribution-v1-001", "prompt": "Closed my entire copper position this morning. I'd rather sit in cash than fight this tape.", "expected_label": "explicit_position", "predicted_label": "quoted", "latency_ms": 20.8, "scores": { "conditional": 0.17577002942562103, "explicit_position": 0.01543655525892973, "news": 0.14686352014541626, "promotion": 0.019054559990763664, "quoted": 0.5839623808860779, "unclear": 0.058913007378578186 }, "correct": false }, { "model_id": "gliclass", "case_id": "social-attribution-v1-002", "prompt": "My take: the regional banks are uninvestable right now and I won't touch them.", "expected_label": "explicit_position", "predicted_label": "quoted", "latency_ms": 22.26, "scores": { "conditional": 0.11087798327207565, "explicit_position": 0.06348104774951935, "news": 0.07206949591636658, "promotion": 0.0024819292593747377, "quoted": 0.7353026866912842, "unclear": 0.0157867930829525 }, "correct": false }, { "model_id": "gliclass", "case_id": "social-attribution-v1-003", "prompt": "I bought the dip in the semis basket today. Full size, no hedge, and I'm comfortable holding through earnings.", "expected_label": "explicit_position", "predicted_label": "quoted", "latency_ms": 22.45, "scores": { "conditional": 0.06324335932731628, "explicit_position": 0.30005306005477905, "news": 0.1723165512084961, "promotion": 0.011842761188745499, "quoted": 0.4330017566680908, "unclear": 0.019542522728443146 }, "correct": false }, { "model_id": "gliclass", "case_id": "social-attribution-v1-004", "prompt": "Staying overweight energy. I've said it for weeks and nothing in this print changes my mind.", "expected_label": "explicit_position", "predicted_label": "quoted", "latency_ms": 23.43, "scores": { "conditional": 0.06725101917982101, "explicit_position": 0.008441722951829433, "news": 0.1604725569486618, "promotion": 0.005331156775355339, "quoted": 0.7374086976051331, "unclear": 0.021094808354973793 }, "correct": false }, { "model_id": "gliclass", "case_id": "social-attribution-v1-005", "prompt": "I trimmed half of my gold miners into the rip. Locking gains, still hold a starter.", "expected_label": "explicit_position", "predicted_label": "quoted", "latency_ms": 23.39, "scores": { "conditional": 0.03940736874938011, "explicit_position": 0.092379130423069, "news": 0.07601305842399597, "promotion": 0.009474541060626507, "quoted": 0.7777289152145386, "unclear": 0.004997034091502428 }, "correct": false }, { "model_id": "gliclass", "case_id": "social-attribution-v1-006", "prompt": "Short the homebuilders. That's where my book is positioned and I'm sizing it up into the meeting.", "expected_label": "explicit_position", "predicted_label": "quoted", "latency_ms": 25.37, "scores": { "conditional": 0.0888189971446991, "explicit_position": 0.05005110800266266, "news": 0.02051853947341442, "promotion": 0.03445535525679588, "quoted": 0.7807631492614746, "unclear": 0.02539285644888878 }, "correct": false }, { "model_id": "gliclass", "case_id": "social-attribution-v1-007", "prompt": "I'm flat everything into the number. No positions, no exposure, watching from the sideline by choice.", "expected_label": "explicit_position", "predicted_label": "news", "latency_ms": 23.63, "scores": { "conditional": 0.07352083921432495, "explicit_position": 0.0020219944417476654, "news": 0.5310638546943665, "promotion": 0.013169101439416409, "quoted": 0.33021071553230286, "unclear": 0.05001349747180939 }, "correct": false }, { "model_id": "gliclass", "case_id": "social-attribution-v1-008", "prompt": "Been accumulating the uranium names all month and I'll keep buying weakness. Conviction is high.", "expected_label": "explicit_position", "predicted_label": "quoted", "latency_ms": 24.77, "scores": { "conditional": 0.11174501478672028, "explicit_position": 0.06696881353855133, "news": 0.01653093285858631, "promotion": 0.09213894605636597, "quoted": 0.6805180907249451, "unclear": 0.03209825977683067 }, "correct": false }, { "model_id": "gliclass", "case_id": "social-attribution-v1-009", "prompt": "I sold my airlines yesterday and I'm not looking back. Wrong thesis, cut it, moving on.", "expected_label": "explicit_position", "predicted_label": "quoted", "latency_ms": 24.14, "scores": { "conditional": 0.19182883203029633, "explicit_position": 0.08167251199483871, "news": 0.09639191627502441, "promotion": 0.016217797994613647, "quoted": 0.568936288356781, "unclear": 0.04495267570018768 }, "correct": false }, { "model_id": "gliclass", "case_id": "social-attribution-v1-010", "prompt": "If VLR reclaims the 200-day on volume, I'll start a position. Until then I do nothing.", "expected_label": "conditional", "predicted_label": "quoted", "latency_ms": 23.42, "scores": { "conditional": 0.3502935767173767, "explicit_position": 0.00159454345703125, "news": 0.22634021937847137, "promotion": 0.030543917790055275, "quoted": 0.37972790002822876, "unclear": 0.011499812826514244 }, "correct": false }, { "model_id": "gliclass", "case_id": "social-attribution-v1-011", "prompt": "Should the jobs number come in soft, I'd add to my rate-cut trades. It hasn't yet, so no move.", "expected_label": "conditional", "predicted_label": "conditional", "latency_ms": 23.58, "scores": { "conditional": 0.47753819823265076, "explicit_position": 0.0008587592164985836, "news": 0.3241509199142456, "promotion": 0.004095267970114946, "quoted": 0.17989112436771393, "unclear": 0.013465767726302147 }, "correct": true }, { "model_id": "gliclass", "case_id": "social-attribution-v1-012", "prompt": "As long as crude holds above seventy, the tanker thesis stays alive for me. Lose it and I'm out.", "expected_label": "conditional", "predicted_label": "quoted", "latency_ms": 25.44, "scores": { "conditional": 0.11403626203536987, "explicit_position": 0.17801496386528015, "news": 0.03153616189956665, "promotion": 0.01268415804952383, "quoted": 0.6540442705154419, "unclear": 0.009684172458946705 }, "correct": false }, { "model_id": "gliclass", "case_id": "social-attribution-v1-013", "prompt": "I'll get bullish the moment this downtrend line breaks. Not before, not on hope.", "expected_label": "conditional", "predicted_label": "quoted", "latency_ms": 25.55, "scores": { "conditional": 0.4549531936645508, "explicit_position": 0.011236004531383514, "news": 0.05611743405461311, "promotion": 0.005997943691909313, "quoted": 0.45562487840652466, "unclear": 0.016070503741502762 }, "correct": false }, { "model_id": "gliclass", "case_id": "social-attribution-v1-014", "prompt": "When the VIX finally cracks under fifteen I'll redeploy the cash pile. We're not there.", "expected_label": "conditional", "predicted_label": "conditional", "latency_ms": 23.48, "scores": { "conditional": 0.6755592226982117, "explicit_position": 0.0004948731511831284, "news": 0.06587180495262146, "promotion": 0.010146965272724628, "quoted": 0.21177856624126434, "unclear": 0.03614850342273712 }, "correct": true }, { "model_id": "gliclass", "case_id": "social-attribution-v1-015", "prompt": "Provided the guidance isn't cut, I stay in. A cut and I reassess the whole position overnight.", "expected_label": "conditional", "predicted_label": "quoted", "latency_ms": 27.16, "scores": { "conditional": 0.03571465611457825, "explicit_position": 0.005723611451685429, "news": 0.014045655727386475, "promotion": 0.012147621251642704, "quoted": 0.9275462031364441, "unclear": 0.004822299815714359 }, "correct": false }, { "model_id": "gliclass", "case_id": "social-attribution-v1-016", "prompt": "If they lose that support shelf I flip short in a heartbeat. Watching it tick by tick.", "expected_label": "conditional", "predicted_label": "quoted", "latency_ms": 25.94, "scores": { "conditional": 0.032623715698719025, "explicit_position": 0.0042064376175403595, "news": 0.05990694463253021, "promotion": 0.01548229530453682, "quoted": 0.8840863108634949, "unclear": 0.0036943417508155107 }, "correct": false }, { "model_id": "gliclass", "case_id": "social-attribution-v1-017", "prompt": "Only if the spread tightens back under two hundred would I touch the high-yield names again.", "expected_label": "conditional", "predicted_label": "quoted", "latency_ms": 23.19, "scores": { "conditional": 0.445923388004303, "explicit_position": 0.0015567524824291468, "news": 0.04597624018788338, "promotion": 0.03240463137626648, "quoted": 0.4646330773830414, "unclear": 0.00950587261468172 }, "correct": false }, { "model_id": "gliclass", "case_id": "social-attribution-v1-018", "prompt": "Assuming the merger clears review, the arb is worth it. If it breaks, this trade is dead money.", "expected_label": "conditional", "predicted_label": "conditional", "latency_ms": 23.43, "scores": { "conditional": 0.45946767926216125, "explicit_position": 0.006481995340436697, "news": 0.08642814308404922, "promotion": 0.001585808233357966, "quoted": 0.4147084951400757, "unclear": 0.031327925622463226 }, "correct": true }, { "model_id": "gliclass", "case_id": "social-attribution-v1-019", "prompt": "I'd chase the breakout, but only on a daily close above the prior high. Intraday pokes don't count.", "expected_label": "conditional", "predicted_label": "quoted", "latency_ms": 23.48, "scores": { "conditional": 0.3778627812862396, "explicit_position": 0.012016058899462223, "news": 0.09564664214849472, "promotion": 0.019834861159324646, "quoted": 0.46590456366539, "unclear": 0.02873520739376545 }, "correct": false }, { "model_id": "gliclass", "case_id": "social-attribution-v1-020", "prompt": "The central bank left rates unchanged and signaled one more hike is possible before year-end.", "expected_label": "news", "predicted_label": "quoted", "latency_ms": 26.58, "scores": { "conditional": 0.02252747118473053, "explicit_position": 0.007004809100180864, "news": 0.2791357934474945, "promotion": 0.0021767104044556618, "quoted": 0.682915985584259, "unclear": 0.006239285692572594 }, "correct": false }, { "model_id": "gliclass", "case_id": "social-attribution-v1-021", "prompt": "Quarterly revenue came in at 4.2 billion, up nine percent from a year ago, the company reported.", "expected_label": "news", "predicted_label": "news", "latency_ms": 23.57, "scores": { "conditional": 0.02864033170044422, "explicit_position": 0.015145833604037762, "news": 0.5194265246391296, "promotion": 0.01962464675307274, "quoted": 0.3952537775039673, "unclear": 0.0219088364392519 }, "correct": true }, { "model_id": "gliclass", "case_id": "social-attribution-v1-022", "prompt": "Regulators approved the acquisition this morning, clearing the last major hurdle for the deal.", "expected_label": "news", "predicted_label": "quoted", "latency_ms": 23.56, "scores": { "conditional": 0.12205899506807327, "explicit_position": 0.02410045452415943, "news": 0.3032640814781189, "promotion": 0.04578576236963272, "quoted": 0.4575496017932892, "unclear": 0.0472412109375 }, "correct": false }, { "model_id": "gliclass", "case_id": "social-attribution-v1-023", "prompt": "Oil settled two percent lower after inventories rose more than analysts had expected.", "expected_label": "news", "predicted_label": "quoted", "latency_ms": 28.75, "scores": { "conditional": 0.02522420324385166, "explicit_position": 0.019963162019848824, "news": 0.26303401589393616, "promotion": 0.0019700282718986273, "quoted": 0.6786575317382812, "unclear": 0.011151080951094627 }, "correct": false }, { "model_id": "gliclass", "case_id": "social-attribution-v1-024", "prompt": "The index closed at a record high, its fourth straight gain, on broad participation across sectors.", "expected_label": "news", "predicted_label": "quoted", "latency_ms": 24.23, "scores": { "conditional": 0.01427432894706726, "explicit_position": 0.008696956560015678, "news": 0.1050766110420227, "promotion": 0.009577552787959576, "quoted": 0.8524380326271057, "unclear": 0.009936561807990074 }, "correct": false }, { "model_id": "gliclass", "case_id": "social-attribution-v1-025", "prompt": "The company announced a recall of two of its product lines pending a safety review.", "expected_label": "news", "predicted_label": "quoted", "latency_ms": 23.47, "scores": { "conditional": 0.11474741250276566, "explicit_position": 0.020578831434249878, "news": 0.3390686810016632, "promotion": 0.02862415835261345, "quoted": 0.4249444007873535, "unclear": 0.07203643769025803 }, "correct": false }, { "model_id": "gliclass", "case_id": "social-attribution-v1-026", "prompt": "Payrolls grew by 180,000 last month and the unemployment rate ticked up to 4.1 percent.", "expected_label": "news", "predicted_label": "quoted", "latency_ms": 23.6, "scores": { "conditional": 0.03242989256978035, "explicit_position": 0.010273769497871399, "news": 0.2549431324005127, "promotion": 0.0032267787028104067, "quoted": 0.6896759867668152, "unclear": 0.00945060234516859 }, "correct": false }, { "model_id": "gliclass", "case_id": "social-attribution-v1-027", "prompt": "A major ratings agency downgraded the sovereign one notch, citing rising debt-service costs.", "expected_label": "news", "predicted_label": "quoted", "latency_ms": 26.74, "scores": { "conditional": 0.05364822596311569, "explicit_position": 0.009953750297427177, "news": 0.23586180806159973, "promotion": 0.005185815040022135, "quoted": 0.6732540130615234, "unclear": 0.022096339613199234 }, "correct": false }, { "model_id": "gliclass", "case_id": "social-attribution-v1-028", "prompt": "The chipmaker said it will build a new fabrication plant, its largest capital outlay to date.", "expected_label": "news", "predicted_label": "quoted", "latency_ms": 23.55, "scores": { "conditional": 0.05078360438346863, "explicit_position": 0.01272773090749979, "news": 0.27237680554389954, "promotion": 0.005316426511853933, "quoted": 0.6418390274047852, "unclear": 0.01695638708770275 }, "correct": false }, { "model_id": "gliclass", "case_id": "social-attribution-v1-029", "prompt": "Freight volumes fell for a third consecutive month, according to the industry group's latest data.", "expected_label": "news", "predicted_label": "quoted", "latency_ms": 23.55, "scores": { "conditional": 0.026517432183027267, "explicit_position": 0.018390078097581863, "news": 0.34740155935287476, "promotion": 0.009523451328277588, "quoted": 0.5686995387077332, "unclear": 0.029467960819602013 }, "correct": false }, { "model_id": "gliclass", "case_id": "social-attribution-v1-030", "prompt": "A viewer asks whether I think lithium has bottomed. Good question, and honestly I'm not the one to call it.", "expected_label": "quoted", "predicted_label": "quoted", "latency_ms": 23.81, "scores": { "conditional": 0.09597384929656982, "explicit_position": 0.00305724935606122, "news": 0.06394641846418381, "promotion": 0.004379104822874069, "quoted": 0.8182002902030945, "unclear": 0.01444315817207098 }, "correct": true }, { "model_id": "gliclass", "case_id": "social-attribution-v1-031", "prompt": "The strategist on the panel argued small caps outperform from here. That's her view, not a claim of mine.", "expected_label": "quoted", "predicted_label": "quoted", "latency_ms": 23.42, "scores": { "conditional": 0.048749469220638275, "explicit_position": 0.01249776966869831, "news": 0.1586216390132904, "promotion": 0.006452002562582493, "quoted": 0.753333568572998, "unclear": 0.020345529541373253 }, "correct": true }, { "model_id": "gliclass", "case_id": "social-attribution-v1-032", "prompt": "Someone in the chat swears this name doubles by spring. I'm just passing along what they said.", "expected_label": "quoted", "predicted_label": "quoted", "latency_ms": 23.54, "scores": { "conditional": 0.05043739080429077, "explicit_position": 0.012155025266110897, "news": 0.31474781036376953, "promotion": 0.006540960632264614, "quoted": 0.5742130279541016, "unclear": 0.041905783116817474 }, "correct": true }, { "model_id": "gliclass", "case_id": "social-attribution-v1-033", "prompt": "My mentor always told me to fade the first gap. Sharing his rule, do with it what you will.", "expected_label": "quoted", "predicted_label": "quoted", "latency_ms": 23.93, "scores": { "conditional": 0.39065274596214294, "explicit_position": 0.07395021617412567, "news": 0.05103819817304611, "promotion": 0.02809249609708786, "quoted": 0.4271981418132782, "unclear": 0.029068203642964363 }, "correct": true }, { "model_id": "gliclass", "case_id": "social-attribution-v1-034", "prompt": "A subscriber emailed that they're loading up on the utilities. Relaying it for the room to weigh.", "expected_label": "quoted", "predicted_label": "quoted", "latency_ms": 23.51, "scores": { "conditional": 0.05088191479444504, "explicit_position": 0.011036211624741554, "news": 0.10743498057126999, "promotion": 0.05366026610136032, "quoted": 0.7669424414634705, "unclear": 0.01004406251013279 }, "correct": true }, { "model_id": "gliclass", "case_id": "social-attribution-v1-035", "prompt": "The famous investor said on stage he'd never short a dull market. Quoting him, not endorsing it.", "expected_label": "quoted", "predicted_label": "quoted", "latency_ms": 28.26, "scores": { "conditional": 0.15607792139053345, "explicit_position": 0.02343102917075157, "news": 0.1019216850399971, "promotion": 0.0243548471480608, "quoted": 0.6101160645484924, "unclear": 0.08409849554300308 }, "correct": true }, { "model_id": "gliclass", "case_id": "social-attribution-v1-036", "prompt": "One of you asked if the dividend is safe. I put the question to the group rather than answer blind.", "expected_label": "quoted", "predicted_label": "quoted", "latency_ms": 26.3, "scores": { "conditional": 0.02734745480120182, "explicit_position": 0.001221063663251698, "news": 0.049511294811964035, "promotion": 0.005918326787650585, "quoted": 0.9103967547416687, "unclear": 0.005604973528534174 }, "correct": true }, { "model_id": "gliclass", "case_id": "social-attribution-v1-037", "prompt": "A caller claimed the sector's already priced the recession. That's his framing; I'm reporting it.", "expected_label": "quoted", "predicted_label": "news", "latency_ms": 23.57, "scores": { "conditional": 0.06067492440342903, "explicit_position": 0.005569822620600462, "news": 0.48468080163002014, "promotion": 0.0033757961355149746, "quoted": 0.4208357632160187, "unclear": 0.02486286498606205 }, "correct": false }, { "model_id": "gliclass", "case_id": "social-attribution-v1-038", "prompt": "The analyst note making the rounds says fair value is ninety. Circulating it, form your own view.", "expected_label": "quoted", "predicted_label": "quoted", "latency_ms": 23.3, "scores": { "conditional": 0.0556829497218132, "explicit_position": 0.026722919195890427, "news": 0.14378686249256134, "promotion": 0.007580701727420092, "quoted": 0.7561178803443909, "unclear": 0.01010868325829506 }, "correct": true }, { "model_id": "gliclass", "case_id": "social-attribution-v1-039", "prompt": "A follower DM'd that insiders are buying. I can't verify it, I'm only surfacing the tip.", "expected_label": "quoted", "predicted_label": "quoted", "latency_ms": 23.86, "scores": { "conditional": 0.07253877073526382, "explicit_position": 0.001642651972360909, "news": 0.1063648983836174, "promotion": 0.04050608351826668, "quoted": 0.7468327879905701, "unclear": 0.032114773988723755 }, "correct": true }, { "model_id": "gliclass", "case_id": "social-attribution-v1-040", "prompt": "Doors to the premium alerts room close Friday. Lock in the founder rate before it goes up.", "expected_label": "promotion", "predicted_label": "quoted", "latency_ms": 25.19, "scores": { "conditional": 0.15195232629776, "explicit_position": 0.010800790973007679, "news": 0.09720960259437561, "promotion": 0.1403307318687439, "quoted": 0.5712365508079529, "unclear": 0.028469989076256752 }, "correct": false }, { "model_id": "gliclass", "case_id": "social-attribution-v1-041", "prompt": "Grab my free options-flow checklist at the link. Drops straight to your inbox, no strings.", "expected_label": "promotion", "predicted_label": "quoted", "latency_ms": 23.48, "scores": { "conditional": 0.15601955354213715, "explicit_position": 0.024644669145345688, "news": 0.16155458986759186, "promotion": 0.14887435734272003, "quoted": 0.38508161902427673, "unclear": 0.12382522225379944 }, "correct": false }, { "model_id": "gliclass", "case_id": "social-attribution-v1-042", "prompt": "This breakdown is brought to you by my charting platform. Use code EDGE for two months free.", "expected_label": "promotion", "predicted_label": "quoted", "latency_ms": 23.49, "scores": { "conditional": 0.07670053839683533, "explicit_position": 0.009382929652929306, "news": 0.3887079656124115, "promotion": 0.07540037482976913, "quoted": 0.4330406188964844, "unclear": 0.016767553985118866 }, "correct": false }, { "model_id": "gliclass", "case_id": "social-attribution-v1-043", "prompt": "Join the weekend workshop where I walk the full playbook live. Seats are limited, link below.", "expected_label": "promotion", "predicted_label": "quoted", "latency_ms": 26.5, "scores": { "conditional": 0.09822612255811691, "explicit_position": 0.029939990490674973, "news": 0.11204953491687775, "promotion": 0.018452512100338936, "quoted": 0.7103166580200195, "unclear": 0.031015237793326378 }, "correct": false }, { "model_id": "gliclass", "case_id": "social-attribution-v1-044", "prompt": "My new course on position sizing is live today. Early birds get the bonus module at no cost.", "expected_label": "promotion", "predicted_label": "quoted", "latency_ms": 23.59, "scores": { "conditional": 0.0557817742228508, "explicit_position": 0.014913062565028667, "news": 0.07669628411531448, "promotion": 0.15480835735797882, "quoted": 0.6874994039535522, "unclear": 0.010301114059984684 }, "correct": false }, { "model_id": "gliclass", "case_id": "social-attribution-v1-045", "prompt": "Subscribe to the pro tier for the full model and the daily watchlist. Cancel anytime, obviously.", "expected_label": "promotion", "predicted_label": "quoted", "latency_ms": 27.49, "scores": { "conditional": 0.12725448608398438, "explicit_position": 0.021175425499677658, "news": 0.044851332902908325, "promotion": 0.3357497751712799, "quoted": 0.3964006304740906, "unclear": 0.07456838339567184 }, "correct": false }, { "model_id": "gliclass", "case_id": "social-attribution-v1-046", "prompt": "Sponsored: open an account with the broker in my bio and get commission-free trades for a year.", "expected_label": "promotion", "predicted_label": "quoted", "latency_ms": 27.11, "scores": { "conditional": 0.1045432910323143, "explicit_position": 0.021624714136123657, "news": 0.09102725237607956, "promotion": 0.11219722777605057, "quoted": 0.653819739818573, "unclear": 0.01678778976202011 }, "correct": false }, { "model_id": "gliclass", "case_id": "social-attribution-v1-047", "prompt": "Want the spreadsheet I use? It's in the members' vault. Membership link is pinned up top.", "expected_label": "promotion", "predicted_label": "quoted", "latency_ms": 24.61, "scores": { "conditional": 0.19907233119010925, "explicit_position": 0.02212982065975666, "news": 0.28331199288368225, "promotion": 0.035670071840286255, "quoted": 0.4360713064670563, "unclear": 0.02374453656375408 }, "correct": false }, { "model_id": "gliclass", "case_id": "social-attribution-v1-048", "prompt": "Last call on the mentorship cohort. Apply tonight and I'll review your book one-on-one.", "expected_label": "promotion", "predicted_label": "quoted", "latency_ms": 24.82, "scores": { "conditional": 0.05766340345144272, "explicit_position": 0.015976229682564735, "news": 0.04938899353146553, "promotion": 0.019371194764971733, "quoted": 0.848770022392273, "unclear": 0.008830180391669273 }, "correct": false }, { "model_id": "gliclass", "case_id": "social-attribution-v1-049", "prompt": "Download the app and follow my live portfolio in real time. First month is on the house.", "expected_label": "promotion", "predicted_label": "quoted", "latency_ms": 23.42, "scores": { "conditional": 0.029382333159446716, "explicit_position": 0.007042999379336834, "news": 0.09882616996765137, "promotion": 0.020747952163219452, "quoted": 0.8377283811569214, "unclear": 0.006272141821682453 }, "correct": false }, { "model_id": "gliclass", "case_id": "social-attribution-v1-050", "prompt": "Interesting setup on this one. We'll see how it plays out.", "expected_label": "unclear", "predicted_label": "quoted", "latency_ms": 24.7, "scores": { "conditional": 0.020395789295434952, "explicit_position": 0.0036350185982882977, "news": 0.03554626926779747, "promotion": 0.0142228864133358, "quoted": 0.9231904745101929, "unclear": 0.003009568201377988 }, "correct": false }, { "model_id": "gliclass", "case_id": "social-attribution-v1-051", "prompt": "Lots of chatter about the sector today. Hard to say which way it breaks from here.", "expected_label": "unclear", "predicted_label": "quoted", "latency_ms": 23.44, "scores": { "conditional": 0.0392645038664341, "explicit_position": 0.002588920295238495, "news": 0.0653887465596199, "promotion": 0.005437708925455809, "quoted": 0.8779981732368469, "unclear": 0.009321940131485462 }, "correct": false }, { "model_id": "gliclass", "case_id": "social-attribution-v1-052", "prompt": "Well, that was a day. On to tomorrow.", "expected_label": "unclear", "predicted_label": "quoted", "latency_ms": 28.78, "scores": { "conditional": 0.12139195203781128, "explicit_position": 0.007800336927175522, "news": 0.11963172256946564, "promotion": 0.01705656759440899, "quoted": 0.7220303416252136, "unclear": 0.01208900474011898 }, "correct": false }, { "model_id": "gliclass", "case_id": "social-attribution-v1-053", "prompt": "Something feels off about this whole move, but I can't put my finger on it yet.", "expected_label": "unclear", "predicted_label": "quoted", "latency_ms": 25.1, "scores": { "conditional": 0.24632969498634338, "explicit_position": 0.0012181071797385812, "news": 0.06177055835723877, "promotion": 0.015704356133937836, "quoted": 0.6572797894477844, "unclear": 0.017697472125291824 }, "correct": false }, { "model_id": "gliclass", "case_id": "social-attribution-v1-054", "prompt": "Charts, charts everywhere. Anyway, hope everyone had a decent session.", "expected_label": "unclear", "predicted_label": "quoted", "latency_ms": 23.16, "scores": { "conditional": 0.0384131483733654, "explicit_position": 0.007034765090793371, "news": 0.12706026434898376, "promotion": 0.00629696249961853, "quoted": 0.8163880109786987, "unclear": 0.004806795623153448 }, "correct": false }, { "model_id": "gliclass", "case_id": "social-attribution-v1-055", "prompt": "Not sure what to make of the print. Could go either way, genuinely.", "expected_label": "unclear", "predicted_label": "quoted", "latency_ms": 24.68, "scores": { "conditional": 0.03463711217045784, "explicit_position": 0.005821156315505505, "news": 0.15549695491790771, "promotion": 0.01569831557571888, "quoted": 0.7753238081932068, "unclear": 0.013022679835557938 }, "correct": false }, { "model_id": "gliclass", "case_id": "social-attribution-v1-056", "prompt": "A lot of moving parts here. Let it settle before deciding anything.", "expected_label": "unclear", "predicted_label": "quoted", "latency_ms": 23.74, "scores": { "conditional": 0.12199865281581879, "explicit_position": 0.005022046621888876, "news": 0.18898405134677887, "promotion": 0.01629333384335041, "quoted": 0.6498875617980957, "unclear": 0.01781441643834114 }, "correct": false }, { "model_id": "gliclass", "case_id": "social-attribution-v1-057", "prompt": "Mixed signals across the board today. No strong read on it.", "expected_label": "unclear", "predicted_label": "quoted", "latency_ms": 23.42, "scores": { "conditional": 0.08043691515922546, "explicit_position": 0.003708391450345516, "news": 0.3226518929004669, "promotion": 0.007485454436391592, "quoted": 0.5682050585746765, "unclear": 0.017512256279587746 }, "correct": false }, { "model_id": "gliclass", "case_id": "social-attribution-v1-058", "prompt": "That headline is either huge or nothing. Time will tell, I suppose.", "expected_label": "unclear", "predicted_label": "quoted", "latency_ms": 26.63, "scores": { "conditional": 0.12196425348520279, "explicit_position": 0.005193029064685106, "news": 0.2699544131755829, "promotion": 0.01387846004217863, "quoted": 0.5663551092147827, "unclear": 0.02265484444797039 }, "correct": false }, { "model_id": "gliclass", "case_id": "social-attribution-v1-059", "prompt": "Watching, thinking, sipping coffee. The market does what it wants.", "expected_label": "unclear", "predicted_label": "quoted", "latency_ms": 27.28, "scores": { "conditional": 0.02749890647828579, "explicit_position": 0.0034033923875540495, "news": 0.13684839010238647, "promotion": 0.029580403119325638, "quoted": 0.7967878580093384, "unclear": 0.00588110089302063 }, "correct": false } ] }