SandbagEval
Benchmark for detecting strategic underperformance in AI models.
Benchmarks
Open benchmarks for AI evaluation. Coming Q2 2026.
Coming Soon
TrustBench
Comprehensive evaluation of AI system trustworthiness.
AgentTrust
Benchmark for multi-agent system trust properties.
Available Now
DistSynth-50
50 distributed systems synthesis tasks for evaluating verified code generation.