Unverified50% confidenceBenchmarkExact time
METR's assessment included Model 4, an unreleased internal model from Anthropic, which exceeded 50% on all 15 tests and exceeded 90% on several
1
Sources
50%
Confidence
Medium-term (~90 days)
Relevance
11/8/2026
First Seen
Valid until: 2/6/2027
Sources
GPT-6 Cyber Limited Release, Gemini Video Editing Upgrade, METR Reports Anthropic R&D Acceleration
bilibili阿梨Aria早鸟报11/6/2026
Related Entities
Related Claims
UnverifiedDuring testing, cases emerged where a model claimed 95% confidence in a move but scored 0 points according to the evaluation criteria.69% similarUnverifiedIn the discovery phase, 15 models were tested; the strongest-performing models were able to discover over 80% of the hidden requirements, while open-weight models could only discover 40-50%.69% similarUnverifiedIn the TRACES test, every model had a failure rate above 71% on the 'agentic' probes; 22 of the 30 models had failure rates above 90%.68% similarUnverifiedOn a held-out test set of 1,200 problems, the calculation scores of 12 models range from 33.3% to 100.0%.66% similarUnverifiedMeta AI research shows that having a model iteratively optimize its own prompts can improve performance by 15-30% on certain standard benchmarks, but carries approximately a 20% probability of falling into a degenerative loop.66% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/1147456API
curl https://kongchang.com/api/v1/knowledge/claims/1147456MCP
get_claim(id=1147456)