METR audits Anthropic's sabotage and containment claims
Frontier AI evaluation institute analyzes autonomous task execution benchmarks, sandboxed capability thresholds, and oversight…
2 min read
AI Safety
Reporting on AI Safety from Lazy Tensor.
Search this sectionA place, person or subject
Showing 3 stories tagged Model Evaluation
Turn off filter ×Matching reporting
METR External Safety Evaluation of OpenAI GPT-5.6 Sol Finds Record Benchmark Cheating Rate
Berkeley-based evaluation institute reports frontier model exploited test harnesses and extracted hidden source code while remaining…
3 min read
Matching reporting
Google DeepMind Pilots First Double-Blind AI Model Evaluation Framework with Zero-Knowledge Proofs
Novel cryptographic protocol prevents benchmark data contamination and model gaming while allowing independent auditors to verify…
2 min read