X post: User: @NielsRogge Post ID: 2097281929124925579 Text: The terrible generalisation of Astra and Fable 5.1 from Terminal Bench 2.1 to 4.0 shows that we’re just training on the
change point: X post: User: @NielsRogge Post ID: 2097281929124925579 Text: The terrible generalisation of Astra and Fable 5.1 from Terminal Bench 2.1 to 4.0 shows that we’re just training on the
This update may affect teams tracking frontier_model_capabilities_and_benchmarks.
{
"event_id": "event_93888dfad7af92c1",
"topic_id": "frontier_model_capabilities_and_benchmarks",
"event_type": "Source Update",
"event_time": "2026-09-08T17:46:43.131023Z",
"title": "NielsRogge:2097281929124925579",
"summary": "X post:\nUser: @NielsRogge\nPost ID: 2097281929124925579\nText: The terrible generalisation of Astra and Fable 5.1 from Terminal Bench 2.1 to 4.0 shows that we’re just training on the",
"contribution": "change point: X post:\nUser: @NielsRogge\nPost ID: 2097281929124925579\nText: The terrible generalisation of Astra and Fable 5.1 from Terminal Bench 2.1 to 4.0 shows that we’re just training on the",
"impact": "This update may affect teams tracking frontier_model_capabilities_and_benchmarks.",
"maturity": "Unknown",
"confidence": 0,
"importance_score": 0.713,
"risk_flags": [
"Unknown Stage"
],
"evidence_count": 1
}Sign in to submit review notes for this event judgment and its evidence trail.