AgentAbstain Benchmark
Explore score history ↗AgentAbstain is a paired-task benchmark for evaluating LLM agents' ability to abstain from acting in scenarios such as ambiguity, conflicting constraints, or tool failures. It includes 263 paired tasks across 42 sandbox environments, with a proposed pipeline for generating fresh task instances.
- Released
- 2026-07-11
- Institutions
- University of Illinois Urbana-Champaign
- Evaluates
- Agents, General AI, Language & Knowledge, Reasoning
- Openness
- unknown
- Importer
- Claire Radar
- Review status
- unreviewed
AgentAbstain paper
No reported scores are on record for this benchmark yet.