paper-with-me

Papers

Beyond Single-Shot: Multi-step Tool Retrieval via Query Planning

2026-01-12 · Wei Fang, James Glass arxiv

LLM agents operating over massive, dynamic tool libraries rely on effective retrieval, yet standard single-shot dense retrievers struggle with complex requests. These failures primarily stem from the disconnect between abstract user goals and technical documentation, and the limited capacity of fixed-size embeddings to model combinatorial tool compositions. To address these challenges, we propose TOOLQP, a lightweight framework that models retrieval as iterative query planning. Instead of single-shot matching, TOOLQP decomposes instructions into sub-tasks and dynamically generates queries to interact with the retriever, effectively bridging the semantic gap by targeting the specific sub-tasks required for composition. We train TOOLQP using synthetic query trajectories followed by optimization via Reinforcement Learning with Verifiable Rewards (RLVR). Experiments demonstrate that TOOLQP achieves state-of-the-art performance, exhibiting superior zero-shot generalization, robustness across diverse retrievers, and significant improvements in downstream agentic execution.

📄 PDF Abstract BibTeX arXiv:2601.07782

Code (0)

등록된 구현이 없습니다.

Tasks

Zero-shot GeneralizationReinforcement Learning

Similar Papers 제목 키워드 기반

ART: Automatic multi-step reasoning and tool-use for large language models

2023-03-16 · Bhargavi Paranjape, Scott Lundberg, Sameer Singh, Hannaneh Hajishirzi 외

Large language models (LLMs) can perform complex reasoning in few- and zero-shot settings by generating intermediate chain of thought (CoT) reasoning steps. Further, each reasoning step can rely on external tools to supp…

MMLU

m&m's: A Benchmark to Evaluate Tool-Use for multi-step multi-modal Tasks

2024-03-17 · Zixian Ma, Weikai Huang, Jieyu Zhang, Tanmay Gupta 외

Real-world multi-modal problems are rarely solved by a single machine learning model, and often require multi-step computational plans that involve stitching several models. Tool-augmented LLMs hold tremendous promise fo…

4k

SPADER: Step-wise Peer Advantage with Diversity-Aware Exploration Rewards for Multi-Answer Question Answering

2026-05-30 · Qiming Shi, Zhaolu Kang, Yunfan Zhou, Di Weng 외 arxiv

Large language models are increasingly deployed as tool-augmented agents to acquire information beyond parametric knowledge. While recent work has improved long-horizon tool-use reasoning, most approaches focus on tasks …

Reinforcement LearningQuestion Answering

HyperTool: Beyond Step-Wise Tool Calls for Tool-Augmented Agents

2026-06-11 · Yaxin Du, Yifan Zhou, Yujie Ge, Jiajun Wang 외 arxiv

Tool-augmented LLM agents commonly rely on step-wise atomic tool calls, where each invocation, observation, and value transfer is exposed in the main reasoning trace. This creates an \emph{execution-granularity mismatch}…

Synthetic Data Generation & Multi-Step RL for Reasoning & Tool Use

2025-04-07 · Anna Goldie, Azalia Mirhoseini, Hao Zhou, Irene Cai 외

Reinforcement learning has been shown to improve the performance of large language models. However, traditional approaches like RLHF or RLAIF treat the problem as single-step. As focus shifts toward more complex reasonin…

GSM8KMathMathematical ReasoningQuestion Answering+4