METR doc confirms Fable 5.2 is 'Model 3 Early', hinting Anthropic holds stronger unreleased models
peterwildeford · x · 2026-11-10
Peter Wildeford shares a METR report screenshot confirming Fable 5.2 is the same snapshot METR evaluated as 'Model 3 Early' in its September pre-deployment assessment, later referenced in METR's October 2026 Independent Assessment of AI R&D Acceleration at Anthropic. This implies Anthropic had more capable internal models before release — likely an unreleased 'Fable 5.5'-class model to match Opus 5.5.
More from Companies & People
- Sequoia investor reveals security benchmark CyScenarioBench was built by portfolio company Irregular — shaunmmaguire · 2026-11-10
- Investor calls WSJ's Nvidia succession-plan story 'journalistic malpractice', says board has a plan — firstadopter · 2026-11-10
- Merge API CTO: your biggest cybersecurity risk is accidentally hiring a spy — shensi · 2026-11-10
- OpenAI's playbook: counter every Anthropic launch within 1-9 days, timeline shows — haider1 · 2026-11-10
- Cerebras releases CS-4 engineering documentary detailing wafer-scale AI system design — draecomino · 2026-11-10
- Semi-Apocalypse podcast: market panic, enterprise AI, and the 2028 GPU glut theory — BenBajarin · 2026-11-10