Community feed
A focused stream of recent stories from the sources curated for this community. Latest: US ramps up frontier AI testing as White House pivots toward safety - MSN, Claude Opus 4.7 Boosts SWE-bench to 87.6% - blockchain.news, and Claude Mythos Shows 50% Time Horizon Of 16+ Hours On METR Benchmark - OfficeChai. Page 40.
Google News Frontier AI Testing
+ 1 source (Google News Frontier AI Testing and Google News AI Testing)
US ramps up frontier AI testing as White House pivots toward safety MSN
Topics: Testing Tools
Entities: Testing Tools
Google News SWE Bench
Claude Opus 4.7 Boosts SWE-bench to 87.6% blockchain.news
Topics: Benchmarks
Entities: ClaudeClaude OpusBenchmarks
Google News AI Safety Evaluation
+ 1 source (Google News AI Safety Evaluation and Google News SWE Bench)
Claude Mythos Shows 50% Time Horizon Of 16+ Hours On METR Benchmark OfficeChai
Topics: Benchmarks
Entities: ClaudeBenchmarksMythos
Google News AI Testing
Amazon's battle to block Perplexity AI testing reach of computer-access laws MLex
Topics: Testing Tools
Entities: Testing Tools
Google News AI Testing
US government AI testing underway - Tech giants to evaluate models Scan TV
Topics: Testing Tools
Entities: Testing Tools
Google News AI Testing
Collate Emphasizes Semantic Data Layer as Key to Enterprise AI Reliability TipRanks
Topics: Testing Tools
Entities: Testing Tools
METR Blog
External review from METR of the "Risks from automated R&D" section in Anthropic's February 2026 Risk Report
Topics: Safety Evals
Entities: AnthropicSafety Evals
METR Blog
We distinguish three measures of AI uplift -- on old tasks, on new tasks, and in value -- and show that task substitution can cause these to diverge substantially.
Google News AI Testing
Tech giants agree to US government AI testing programme Euronews.com
Topics: Testing Tools
Entities: Testing Tools
Google News SWE Bench
Scale Labs debuts new Refactoring Leaderboard for AI TestingCatalog AI News
Topics: Benchmarks
Entities: Benchmarks
Google News MLPerf
Cisco and AMD Benchmark Scale-out AI Fabric Performance Let's Data Science
Topics: Benchmarks
Entities: Benchmarks
MLCommons Evaluation Filter
MLPerf Training v6.0 introduces GPT-OSS 20B, a new sparse Mixture-of-Experts (MoE) pretraining benchmark designed for accessibility on single 8-GPU nodes.
Topics: Benchmarks
Entities: Benchmarks