Spurious Tool Use: When RL Agents Learn the Wrong Reason to Act
Understanding and mitigating shortcut tool selection in RL-trained agents through rewards for tool necessity.
•
1 min read
Read more
Clarify or Answer: Reinforcement Learning for Agentic VQA with Context Under-specification
Reinforcement learning for agentic VQA that balances clarification and answering under underspecified context.
•
1 min read
Read more