Spurious Tool Use: When RL Agents Learn the Wrong Reason to Act

arXiv:2609.16268 · cs.CL · Submitted 2026-09-14 · Read on arXiv

cs.CL

Submitted: 2026-09-14

Updated: 2026-09-14

License: http://arxiv.org/licenses/nonexclusive-distrib/1.0/

Terminology

• Accuracy — 10× in this paper · explained in 1 episode
• Reinforcement Learning — 10× in this paper · explained in 4 episodes
• Alignment — 9× in this paper · explained in 1 episode
• Semantic Alignment — 9× in this paper · explained in 3 episodes
• LLM judge — 7× in this paper · explained in 2 episodes
• Robustness — 7× in this paper · explained in 5 episodes
• Benchmark — 4× in this paper · explained in 2 episodes
• GRPO — 3× in this paper · explained in 6 episodes
• Credit assignment — 2× in this paper · explained in 3 episodes
• Large Language Model (LLM) — 2× in this paper · explained in 3 episodes
• Reinforcement Learning (RL) — 2× in this paper · explained in 25 episodes
• Agentic Search — 1× in this paper · explained in 1 episode
• Agentic System — 1× in this paper · explained in 2 episodes
• Decoupling — 1× in this paper · explained in 2 episodes
• Fine-Tuning — 1× in this paper · explained in 5 episodes
• Generalization — 1× in this paper · explained in 13 episodes
• Large Language Models (LLMs) — 1× in this paper · explained in 6 episodes
• Provenance — 1× in this paper · explained in 4 episodes
• Regularization — 1× in this paper · explained in 3 episodes
• Transformer — 1× in this paper · explained in 2 episodes
• Verifier — 1× in this paper · explained in 2 episodes

Sources

• Invariant Risk Minimization
• ReSearch: Learning to Reason with Search for LLMs via Reinforcement Learning
• Training Verifiers to Solve Math Word Problems
• ReTool: Reinforcement Learning for Strategic Tool Use in LLMs
• DeepMath-103K: A Large-Scale, Challenging, Decontaminated, and Verifiable Mathematical Dataset for Advancing Reasoning
• VerlTool: Towards Holistic Agentic Reinforcement Learning with Tool Use
• Search-R1: Training LLMs to Reason and Leverage Search Engines with Reinforcement Learning
• Last Layer Re-Training is Sufficient for Robustness to Spurious Correlations
• Search-o1: Agentic Search-Enhanced Large Reasoning Models
• ToRL: Scaling Tool-Integrated RL
• In-the-Flow Agentic System Optimization for Effective Planning and Tool Use
• Agent RL Scaling Law: Agent RL with Spontaneous Code Execution for Mathematical Problem Solving
• ART: Automatic multi-step reasoning and tool-use for large language models
• ToolLLM: Facilitating Large Language Models to Master 16000+ Real-world APIs
• Distributionally Robust Neural Networks for Group Shifts: On the Importance of Regularization for Worst-Case Generalization
• DeepSeekMath: Pushing the Limits of Mathematical Reasoning in Open Language Models
• R1-Searcher: Incentivizing the Search Capability in LLMs via Reinforcement Learning
• ZeroSearch: Incentivize the Search Capability of LLMs without Searching
• Reducing Cognitive Overhead in Tool Use via Multi-Small-Agent Reinforcement Learning
• Acting Less is Reasoning More! Teaching Model to Act Efficiently

Related papers

• Exploring Solution Divergence and Its Effect on Large Language Model Problem Solving
• Ishigaki-IDS-Bench: A Benchmark for Generating Information Delivery Specification from BIM Information Requirements
• Subliminal Steering: Stronger Encoding of Hidden Signals
• MedStruct-S: A Benchmark for Key Discovery, Key-Conditioned QA and Semi-Structured Extraction from OCR Clinical Reports
• The End of Transformers? On Challenging Attention and the Rise of Sub-Quadratic Architectures
• Untangling the Mechanisms of Misleading Context in Medical Question Answering