mlsanityData to 7 Oct 2026

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

arXiv 2601.07782 ↗First posted 12 Jan 2026cs.CL cs.AI cs.IRcited by 2

On the map: LLM agents and tool use → Tool learning and function calling

Abstract (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.

Key facts. Cited by 2 papers in the map (data through 7 Oct 2026). Placed in Tool learning and function calling, one of 16 directions in LLM agents and tool use. Builds on 43 of its 76 references that are in the map. First posted on arXiv 12 Jan 2026.

Builds on

43 of 76 references.

Closest ancestors · 12

Same direction, newest first

  1. 2025ToolDreamer: Instilling LLM Reasoning Into Tool Retrievers22 Oct
  2. 2025StableToolBench-MirrorAPI: Modeling Tool Environments as Mirrors of 7,000+ Real-World APIs26 Mar
  3. 2025PLAY2PROMPT: Zero-shot Tool Instruction Optimization for LLM Agents via Tool Play18 Mar
  4. 2024From Exploration to Mastery: Enabling LLMs to Master Tools via Self-Driven Interactions10 Oct
  5. 2024APIGen: Automated Pipeline for Generating Verifiable and Diverse Function-Calling Datasets26 Jun
  6. 2023Confucius: Iterative Tool Learning from Introspection Feedback by Easy-to-Difficult Curriculum27 Aug
  7. 2023ToolLLM: Facilitating Large Language Models to Master 16000+ Real-world APIs31 Jul
  8. 2023GPT4Tools: Teaching Large Language Model to Use Tools via Self-instruction30 May
  9. 2023Large Language Models as Tool Makers26 May
  10. 2023Gorilla: Large Language Model Connected with Massive APIs24 May
  11. 2023API-Bank: A Comprehensive Benchmark for Tool-Augmented LLMs14 Apr
  12. 2023Toolformer: Language Models Can Teach Themselves to Use Tools9 Feb

Other references · 31

  1. 2022ReAct: Synergizing Reasoning and Acting in Language Modelscited by 8,355
  2. 2023Lost in the Middle: How Language Models Use Long Contextscited by 3,383
  3. 2022Text Embeddings by Weakly-Supervised Contrastive Pre-trainingcited by 1,216
  4. 2024M3-Embedding: Multi-Linguality, Multi-Functionality, Multi-Granularity Text Embeddings Through Self-Knowledge Distillationcited by 1,050
  5. 2025Does Reinforcement Learning Really Incentivize Reasoning Capacity in LLMs Beyond the Base Model?cited by 948
  6. 2023Towards General Text Embeddings with Multi-stage Contrastive Learningcited by 705
  7. 2023Visual ChatGPT: Talking, Drawing and Editing with Visual Foundation Modelscited by 659
  8. 2023ViperGPT: Visual Inference via Python Execution for Reasoningcited by 591
  9. 2022Precise Zero-Shot Dense Retrieval without Relevance Labelscited by 508
  10. 2022Visual Programming: Compositional visual reasoning without trainingcited by 507
  11. 2021SPLADE: Sparse Lexical and Expansion Model for First Stage Rankingcited by 283
  12. 2019Document Expansion by Query Predictioncited by 225
19 more
  1. 2021PAQ: 65 Million Probably-Asked Questions and What You Can Do With Themcited by 198
  2. 2020Generation-Augmented Retrieval for Open-domain Question Answeringcited by 186
  3. 2022Improving Passage Retrieval with Zero-Shot Question Generationcited by 164
  4. 2023ToolChain*: Efficient Action Space Navigation in Large Language Models with A* Searchcited by 103
  5. 2025On the Theoretical Limitations of Embedding-Based Retrievalcited by 98
  6. 2020Few-Shot Generative Conversational Query Rewritingcited by 85
  7. 2023Large Language Models Know Your Contextual Search Intent: A Prompting Framework for Conversational Searchcited by 74
  8. 2024ToolGen: Unified Tool Retrieval and Calling via Generationcited by 72
  9. 2023Enhancing Conversational Search: Large Language Model-Aided Informative Query Rewritingcited by 69
  10. 2019Context-Aware Sentence/Passage Term Importance Estimation For First Stage Retrievalcited by 65
  11. 2024Re-Invoke: Tool Invocation Rewriting for Zero-Shot Tool Retrievalcited by 50
  12. 2021Improving Query Representations for Dense Retrieval with Pseudo Relevance Feedbackcited by 45
  13. 2021Pseudo-Relevance Feedback for Multiple Representation Dense Retrievalcited by 31
  14. 2021Pseudo Relevance Feedback with Deep Language Models and Dense Retrievers: Successes and Pitfallscited by 26
  15. 2023Large Language Models are Strong Zero-Shot Retrievercited by 25
  16. 2023Expand, Rerank, and Retrieve: Query Reranking for Open-Domain Question Answeringcited by 15
  17. 2023Natural Language Embedded Programs for Hybrid Language Symbolic Reasoningcited by 15
  18. 2025Tools are under-documented: Simple Document Expansion Boosts Tool Retrievalcited by 12
  19. 2023Synergistic Interplay between Search and Large Language Models for Information Retrievalcited by 11

Built on by

Cited by year (2026 to 7 Oct)

Citing papers per year: 2026: 2.

2 papers.

  1. 2026MagicSelector: Joint Optimization for Agent Tool Selection via Counterfactual Decomposition and Progressive Rerankingcited by 1
  2. 2026Factorized Hypothesis Search for Evidence-to-Taxonomy Retrievalcited by 0

Similar, not cited

Close in wording, no citation either way

  1. 2024Toolshed: Scale Tool-Equipped Agents with Advanced RAG-Tool Fusion and Tool Knowledge Bases
  2. 2026Tools Are Not Islands: Set-Level Tool Retrieval for LLM Agents via Query-Conditioned Hyperedge Prediction
  3. 2024Data-Efficient Massive Tool Retrieval: A Reinforcement Learning Approach for Query-Tool Alignment with Language Models
  4. 2026CoHyDE: Iterative Co-Training of LLM Rewriter & Dense Encoder for Tool Retrieval
  5. 2025Tool-to-Agent Retrieval: Bridging Tools and Agents for Scalable LLM Multi-Agent Systems
  6. 2026ToolSearcher: Optimizing Tool Selection at Scale via Reinforcement Learning
  7. 2026Lookahead-R: Budget-Aware Tool Retrieval via Execution-Centric Planning
  8. 2026SING: Synthetic Intention Graph for Scalable Active Tool Discovery in LLM Agents