UndoBench: Separating Task Competence from Recovery Capability in Tool-Using AI Agents
Paper • 2610.05622 • Published • 13 •
3
Tanmay Sah
TanmaySah
AI & ML interests
AI agents, LLM reasoning, ARC-AGI, verifiers, agent harnesses, recovery, model evaluation, and trustworthy AI
Recent Activity
updated a collection 23 days ago
Self Evolution updated a Space 26 days ago
TanmaySah/InvariantOne-Demo updated a model 26 days ago
TanmaySah/InvariantOne-v1Organizations
None yet
commented a
paper about 1 month ago