None defined yet.
Invariant-Aware Harness Enforcement to Protect LLM Agents from Silent API Failures
AutoSciBench: Autonomous Benchmark Generation for Evaluating Scientific Agents