mlsanityData to 7 Oct 2026

Beyond Static Tools: Test-Time Tool Evolution for Scientific Reasoning

arXiv 2601.07641 ↗First posted 12 Jan 2026cs.AI cs.CL cs.MAcited by 12

On the map: LLM reasoning → Reasoning for agents, planning and code

Abstract (arXiv)

The central challenge of AI for Science is not reasoning alone, but the ability to create computational methods in an open-ended scientific world. Existing LLM-based agents rely on static, pre-defined tool libraries, a paradigm that fundamentally fails in scientific domains where tools are sparse, heterogeneous, and intrinsically incomplete. In this paper, we propose Test-Time Tool Evolution (TTE), a new paradigm that enables agents to synthesize, verify, and evolve executable tools during inference. By transforming tools from fixed resources into problem-driven artifacts, TTE overcomes the rigidity and long-tail limitations of static tool libraries. To facilitate rigorous evaluation, we introduce SciEvo, a benchmark comprising 1,590 scientific reasoning tasks supported by 925 automatically evolved tools. Extensive experiments show that TTE achieves state-of-the-art performance in both accuracy and tool efficiency, while enabling effective cross-domain adaptation of computational tools. The code and benchmark have been released at https://github.com/lujiaxuan0520/Test-Time-Tool-Evol.

Key facts. Cited by 12 papers in the map (data through 7 Oct 2026). Placed in Reasoning for agents, planning and code, one of 17 directions in LLM reasoning. Builds on 24 of its 27 references that are in the map. First posted on arXiv 12 Jan 2026.

Builds on

24 of 27 references.

References · 24

  1. 2020Language Models are Few-Shot Learnerscited by 32,199
  2. 2019Sentence-BERT: Sentence Embeddings using Siamese BERT-Networkscited by 9,049
  3. 2022ReAct: Synergizing Reasoning and Acting in Language Modelscited by 8,355
  4. 2021Evaluating Large Language Models Trained on Codecited by 8,053
  5. 2023Toolformer: Language Models Can Teach Themselves to Use Toolscited by 4,072
  6. 2023ToolLLM: Facilitating Large Language Models to Master 16000+ Real-world APIscited by 1,880
  7. 2023Gorilla: Large Language Model Connected with Massive APIscited by 1,302
  8. 2024M3-Embedding: Multi-Linguality, Multi-Functionality, Multi-Granularity Text Embeddings Through Self-Knowledge Distillationcited by 1,050
  9. 2023ChemCrow: Augmenting large-language models with chemistry toolscited by 609
  10. 2023API-Bank: A Comprehensive Benchmark for Tool-Augmented LLMscited by 574
  11. 2023Large Language Models as Tool Makerscited by 293
  12. 2023CRAFT: Customizing LLMs by Creating and Retrieving from Specialized Toolsetscited by 122
12 more
  1. 2023SciEval: A Multi-Level Large Language Model Evaluation Benchmark for Scientific Researchcited by 112
  2. 2024HumanEval-XL: A Multilingual Code Generation Benchmark for Cross-lingual Natural Language Generalizationcited by 84
  3. 2025SEAgent: Self-Evolving Computer Use Agent with Autonomous Learning from Experiencecited by 68
  4. 2024HoneyComb: A Flexible LLM-Based Agent System for Materials Sciencecited by 56
  5. 2024NESTFUL: A Benchmark for Evaluating LLMs on Nested Sequences of API Callscited by 45
  6. 2024Are LLMs Ready for Real-World Materials Discovery?cited by 40
  7. 2025LongFuncEval: Measuring the effectiveness of long context models for function callingcited by 17
  8. 2025CONFETTI: Conversational Function-Calling Evaluation Through Turn-Level Interactionscited by 16
  9. 2025CheMatAgent: Enhancing LLMs for Chemistry and Materials Science through Tree-Search Based Tool Learningcited by 12
  10. 2025MatTools: Benchmarking Large Language Models for Materials Science Toolscited by 11
  11. 2025From What to Why: A Multi-Agent System for Evidence-based Chemical Reaction Condition Reasoningcited by 10
  12. 2025ToolACE-DEV: Self-Improving Tool Learning via Decomposition and EVolutioncited by 7

Built on by

Cited by year (2026 to 7 Oct)

Citing papers per year: 2026: 12.

12 papers.

  1. 2026CoEvoSkills: Self-Evolving Agent Skills via Co-Evolutionary Verificationcited by 117
  2. 2026SkillsVote: Lifecycle Governance of Agent Skills from Collection, Recommendation to Evolutioncited by 18
  3. 2026The Evolution of Tool Use in LLM Agents: From Single-Tool Call to Multi-Tool Orchestrationcited by 11
  4. 2026FinToolBench: Evaluating LLM Agents for Real-World Financial Tool Usecited by 7
  5. 2026Recursive Self-Evolving Agents via Held-Out Selectioncited by 6
  6. 2026Self-Improvement of Large Language Models: A Technical Overview and Future Outlookcited by 5
  7. 2026The Last AI Built by Humans: Toward Genuine Recursive Self-Improvementcited by 5
  8. 2026Yunjue Agent Tech Report: A Fully Reproducible, Zero-Start In-Situ Self-Evolving Agent System for Open-Ended Taskscited by 4
  9. 2026VESTA: Visual Exploration with Statistical Tool Agentscited by 1
  10. 2026Metacognition in LLMs: Foundations, Progress, and Opportunitiescited by 1
  11. 2026Eta Given Delta: Defining LLM Tool Efficiency With Marginal Tool Utilitycited by 0
  12. 2026Harnessing agent memory to build lifelong AI partners for materials scientistscited by 0

Similar, not cited

Close in wording, no citation either way

  1. 2026SciToolAgent-Evo: An Ontology-Aware Self-Evolving Agent for Open-World Scientific Tool Acquisition
  2. 2026SciAgentGym: Benchmarking Multi-Step Scientific Tool-use in LLM Agents
  3. 2026Evoflux: Inference-Time Evolution of Executable Tool Workflows for Compact Agents
  4. 2026ScienceClaw: Benchmarking Continual Self-Evolution of AI-for-Science Agents Across the Natural and Social Sciences
  5. 2026CliffSearch: Structured Agentic Co-Evolution over Theory and Code for Scientific Algorithm Discovery
  6. 2025Evolution without an Oracle: Driving Effective Evolution with LLM Judges
  7. 2026TTHE: Test-Time Harness Evolution
  8. 2025ToolLibGen: Scalable Automatic Tool Creation and Aggregation for LLM Reasoning