---
title: "ToolScope — paper by Liu et al. (2026) — summary and ideas"
description: "The work offers an automated, scalable remedy for noisy toolsets and shows that public benchmarks… 43 references, cited by 1 paper here."
authors: ["Marianne Menglin Liu", "Daniel Garcia", "Fjona Parllaku", "Vikas Upadhyay", "Syed Fahad Allam Shah", "Dan Roth"]
year: 2026
venue: "Proceedings of the 64th Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers)"
source: "https://aclanthology.org/2026.acl-long.1573/"
doi: "10.18653/v1/2026.acl-long.1573"
licence: "CC BY 4.0 (ACL Anthology)"
licence_url: "https://creativecommons.org/licenses/by/4.0/"
summary_licence: "Elamur Cosmic License"
canonical: "https://research.elamur.ai/free/en/papers/liu-2026-toolscope-enhancing-llm-agent-tool/"
---
[Elamur Research](https://md.research.elamur.ai/free/en/) · [this page in HTML](https://research.elamur.ai/free/en/papers/liu-2026-toolscope-enhancing-llm-agent-tool/)

[home](https://md.research.elamur.ai/free/en/) / [papers](https://md.research.elamur.ai/free/en/papers/) / [2026](https://md.research.elamur.ai/free/en/papers/2026/page/4/) / ToolScope: Enhancing LLM Agent Tool Use through Tool Merging and Context-Aware Filtering

# ToolScope: Enhancing LLM Agent Tool Use through Tool Merging and Context-Aware Filtering

**Authors:** [Marianne Menglin Liu](https://md.research.elamur.ai/free/en/authors/marianne-menglin-liu/), Daniel Garcia, Fjona Parllaku, Vikas Upadhyay, Syed Fahad Allam Shah and [Dan Roth](https://md.research.elamur.ai/free/en/authors/dan-roth/) · **Year:** 2026 · **Venue:** [Proceedings of the 64th Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers)](https://md.research.elamur.ai/free/en/venues/acl/) ([ACL 2026](https://md.research.elamur.ai/free/en/venues/acl/2026/)) · **Publisher:** [Association for Computational Linguistics](https://md.research.elamur.ai/free/en/publishers/association-for-computational-linguistics/) · **Licence:** [CC BY 4.0 (ACL Anthology)](https://creativecommons.org/licenses/by/4.0/)

**Links:** [Publisher's page](https://aclanthology.org/2026.acl-long.1573/) · [PDF](https://aclanthology.org/2026.acl-long.1573.pdf) · [DOI 10.18653/v1/2026.acl-long.1573](https://doi.org/10.18653/v1/2026.acl-long.1573) · [OpenAlex](https://openalex.org/W4416245059)

**Citations:** 3 citations (OpenAlex, 2026-10-14) · 1 in our library, of the 149,439 papers whose bibliographies we read

## Summary
LLM agents often choose the wrong external tool when a toolset contains many near-duplicate tools with similar names and descriptions. Large toolsets also overflow the model's input window. ToolScope (Oracle AI) addresses both problems without fine-tuning.

Its first component, ToolScopeMerger, works in several steps:
- It embeds tool signatures and descriptions and finds candidate neighbours by cosine similarity.
- An LLM classifier judges whether each candidate pair is functionally equivalent.
- Equivalent pairs become edges in a graph, and each connected component becomes a merge cluster.
- An optional Auto-Correction step lets an LLM validator audit and split faulty clusters.
- Merged tools receive LLM-written unified descriptions, and benchmark gold labels are remapped to match.

The second component, ToolScopeRetriever, decomposes queries, mixes sparse (BM25) and dense scores, reranks and normalizes the results, and passes only the top-k tools to the agent.

The authors tested ToolScope with three LLMs on BFCL, Seal-Tools and UltraTool. Tool selection accuracy improved by up to 38.6%. The work offers an automated, scalable remedy for noisy toolsets and shows that public benchmarks contain substantial redundancy.

## Key points
- Selection accuracy gains over baselines: 34.6% on Seal-Tools, 38.6% on UltraTool and 8.8% on BFCL. The abstract reports the lower bound as 8.38%.
- Average context per query shrinks sharply:
  - BFCL: 32,563 → 469 tokens
  - Seal-Tools: 292,107 → 317 tokens
  - UltraTool: 136,352 → 2,076 tokens
  - This amounts to a reduction of roughly 98.5–99.9%.
- Merging raises Recall@10 from 0.550 to 0.935 on Seal-Tools and from 0.945 to 0.985 on BFCL.
- Auto-Correction adds CSR gains of 1.5–2.9% on Seal-Tools and 2.9–7.9% on UltraTool. On BFCL it adds 2.6% with GPT-4o and 5.0% with LlaMa-3.3-70b.
- Additional findings:
  - Results hold even for low-quality tool documentation, as rated by a GPT-4o judge. Low-band CSR is 85.7% (BFCL), 95.0% (Seal-Tools) and 72.7% (UltraTool).
  - Reranking helps most at small k (5–10).
  - Tuning the sparse/dense weight α showed retrieval peaks at α = 1, i.e. dense-only.

*Our summary of the paper, in our own words — not its text. The work belongs to its authors; read it at the links above.*

📚 Shelves (5): [🖥️ Computer Science](https://md.research.elamur.ai/free/en/shelves/computer-science/) · 16,667 works · [🤖 Artificial Intelligence](https://md.research.elamur.ai/free/en/shelves/computer-science/artificial-intelligence/) · 16,194 works · [🎭 Arts and Humanities](https://md.research.elamur.ai/free/en/shelves/arts-and-humanities/) · 16,039 works · [🗣️ Language and Linguistics](https://md.research.elamur.ai/free/en/shelves/arts-and-humanities/language-and-linguistics/) · 15,891 works · [🗄️ Information Systems](https://md.research.elamur.ai/free/en/shelves/computer-science/information-systems/) · 1,437 works

## Cite

Cite as: Marianne Menglin Liu, Daniel Garcia, Fjona Parllaku, Vikas Upadhyay, Syed Fahad Allam Shah and Dan Roth. 2026. ToolScope: Enhancing LLM Agent Tool Use through Tool Merging and Context-Aware Filtering. Proceedings of the 64th Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers).

The paper itself, not this page. BibTeX from the library's record of the work.

```bibtex
@inproceedings{liu2026toolscope,
  title = {{ToolScope: Enhancing LLM Agent Tool Use through Tool Merging and Context-Aware Filtering}},
  author = {Marianne Menglin Liu, Daniel Garcia, Fjona Parllaku, Vikas Upadhyay, Syed Fahad Allam Shah and Dan Roth},
  booktitle = {Proceedings of the 64th Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers)},
  year = {2026},
  doi = {10.18653/v1/2026.acl-long.1573},
  url = {https://doi.org/10.18653/v1/2026.acl-long.1573}
}
```

- [← previous TLSA: LLM-Guided Text-Label Space Alignment with Contrastive Learning for Generalized Category Discovery](https://md.research.elamur.ai/free/en/papers/xu-2026-tlsa-llm-guided-text-label/)
- [next → Traffic-R1: Reinforced LLMs Bring Human-Like Reasoning to Traffic Signal Control Systems](https://md.research.elamur.ai/free/en/papers/zou-2026-traffic-r1-reinforced-llms-bring/)
- [↑ back to papers of 2026](https://md.research.elamur.ai/free/en/papers/2026/page/4/)

## References

43 references, 10 of them with a card of ours. ↗ leads to the work, «find» searches for one we have no id for yet.

1. Emre Can Acikgoz, Jeremiah Greer, Akul Datta, Ze Yang, William Zeng, Oussama Elachqar, Emmanouil Koukoumidis, Dilek Hakkani-Tür, and Gokhan Tur. 2025. Can a single model master both multi-turn conversations and tool use? calm: A unified conversational agentic language model. arXiv preprint arXiv:250… — [arXiv:2502.08820](https://arxiv.org/abs/2502.08820) [find](https://scholar.google.com/scholar?q=Emre+Can+Acikgoz%2C+Jeremiah+Greer%2C+Akul+Datta%2C+Ze+Yang%2C+William+Zeng%2C+Oussama+Elachqar%2C+Emmanouil+Koukoumidis%2C+Dilek+Hakkani%2DT%C3%BCr%2C+and+Gokhan+Tur%2E+2025%2E+Can+a+single+model+master+both+multi%2Dturn+convers)
2. Lisa Alazraki and Marek Rei. 2025. Meta-reasoning improves tool use in large language models. In Findings of the Association for Computational Linguistics: NAACL 2025, page 7885–7897. Association for Computational Linguistics. [find](https://scholar.google.com/scholar?q=Lisa+Alazraki+and+Marek+Rei%2E+2025%2E+Meta%2Dreasoning+improves+tool+use+in+large+language+models%2E+In+Findings+of+the+Association+for+Computational+Linguistics%3A+NAACL+2025%2C+page+7885%E2%80%937897%2E+Association+for+)
3. Nuo Chen, Hongguang Li, Baoyuan Wang, and Jia Li. 2023. From good to great: Improving math reasoning with tool-augmented interleaf prompting. arXiv preprint arXiv:2401.05384. [find](https://scholar.google.com/scholar?q=Nuo+Chen%2C+Hongguang+Li%2C+Baoyuan+Wang%2C+and+Jia+Li%2E+2023%2E+From+good+to+great%3A+Improving+math+reasoning+with+tool%2Daugmented+interleaf+prompting%2E+arXiv+preprint+arXiv%3A2401%2E05384%2E)
4. Yanfei Chen, Jinsung Yoon, Devendra Singh Sachan, Qingze Wang, Vincent Cohen-Addad, Mohammadhossein Bateni, Chen-Yu Lee, and Tomas Pfister. 2024. Re-invoke: Tool invocation rewriting for zero-shot tool retrieval. arXiv preprint arXiv:2408.01875. [find](https://scholar.google.com/scholar?q=Yanfei+Chen%2C+Jinsung+Yoon%2C+Devendra+Singh+Sachan%2C+Qingze+Wang%2C+Vincent+Cohen%2DAddad%2C+Mohammadhossein+Bateni%2C+Chen%2DYu+Lee%2C+and+Tomas+Pfister%2E+2024%2E+Re%2Dinvoke%3A+Tool+invocation+rewriting+for+zero%2Dshot+too)
5. Cohere. 2024. Command models get an august refresh. Accessed: 2025-05-18. [find](https://scholar.google.com/scholar?q=Cohere%2E+2024%2E+Command+models+get+an+august+refresh%2E+Accessed%3A+2025%2D05%2D18%2E)
6. Cross-Encoder. 2024. cross-encoder/ms-marco-minilm-l6-v2. Accessed: 2025-05-18. [find](https://scholar.google.com/scholar?q=Cross%2DEncoder%2E+2024%2E+cross%2Dencoder%2Fms%2Dmarco%2Dminilm%2Dl6%2Dv2%2E+Accessed%3A+2025%2D05%2D18%2E)
7. Matthijs Douze, Alexandr Guzhva, Chengqi Deng, Jeff Johnson, Gergely Szilvasy, Pierre-Emmanuel Mazaré, Maria Lomeli, Lucas Hosseini, and Hervé Jégou. 2024. [The faiss library](https://md.research.elamur.ai/free/en/papers/douze-2024-faiss-library/). arXiv preprint arXiv:2401.08281.
8. Shijue Huang, Wanjun Zhong, Jianqiao Lu, Qi Zhu, Jiahui Gao, Weiwen Liu, Yutai Hou, Xingshan Zeng, Yasheng Wang, Lifeng Shang, Xin Jiang, Ruifeng Xu, and Qun Liu. 2024. Planning, creation, usage: Benchmarking llms for comprehensive tool utilization in real-world complex scenarios. Preprint, arXiv:24… [find](https://scholar.google.com/scholar?q=Shijue+Huang%2C+Wanjun+Zhong%2C+Jianqiao+Lu%2C+Qi+Zhu%2C+Jiahui+Gao%2C+Weiwen+Liu%2C+Yutai+Hou%2C+Xingshan+Zeng%2C+Yasheng+Wang%2C+Lifeng+Shang%2C+Xin+Jiang%2C+Ruifeng+Xu%2C+and+Qun+Liu%2E+2024%2E+Planning%2C+creation%2C+usage%3A+Benc)
9. Yue Huang, Jiawen Shi, Yuan Li, Chenrui Fan, Siyuan Wu, Qihui Zhang, Yixin Liu, Pan Zhou, Yao Wan, Neil Zhenqiang Gong, and 1 others. 2023. Meta-tool benchmark for large language models: Deciding whether to use tools and which to use. arXiv preprint arXiv:2310.03128. [find](https://scholar.google.com/scholar?q=Yue+Huang%2C+Jiawen+Shi%2C+Yuan+Li%2C+Chenrui+Fan%2C+Siyuan+Wu%2C+Qihui+Zhang%2C+Yixin+Liu%2C+Pan+Zhou%2C+Yao+Wan%2C+Neil+Zhenqiang+Gong%2C+and+1+others%2E+2023%2E+Meta%2Dtool+benchmark+for+large+language+models%3A+Deciding+whet)
10. Hugging Face. 2024. smolagents. [https://github.com/huggingface/smolagents](https://github.com/huggingface/smolagents). 34103 9 [↗](https://github.com/huggingface/smolagents)
11. Tatsuro Inaba, Hirokazu Kiyomaru, Fei Cheng, and Sadao Kurohashi. 2023. Multitool-cot: Gpt-3 can use multiple external tools with chain of thought prompting. arXiv preprint arXiv:2305.16896. [find](https://scholar.google.com/scholar?q=Tatsuro+Inaba%2C+Hirokazu+Kiyomaru%2C+Fei+Cheng%2C+and+Sadao+Kurohashi%2E+2023%2E+Multitool%2Dcot%3A+Gpt%2D3+can+use+multiple+external+tools+with+chain+of+thought+prompting%2E+arXiv+preprint+arXiv%3A2305%2E16896%2E)
12. Cheng Qian, Emre Can Acikgoz, Qi He, Hongru Wang, Xiusi Chen, Dilek Hakkani-Tür, Gokhan Tur, and Heng Ji. 2025. Toolrl: Reward is all tool learning needs. arXiv preprint arXiv:2504.13958. [find](https://scholar.google.com/scholar?q=Cheng+Qian%2C+Emre+Can+Acikgoz%2C+Qi+He%2C+Hongru+Wang%2C+Xiusi+Chen%2C+Dilek+Hakkani%2DT%C3%BCr%2C+Gokhan+Tur%2C+and+Heng+Ji%2E+2025%2E+Toolrl%3A+Reward+is+all+tool+learning+needs%2E+arXiv+preprint+arXiv%3A2504%2E13958%2E)
13. Bowen Jin, Hansi Zeng, Zhenrui Yue, Jinsung Yoon, Sercan Arik, Dong Wang, Hamed Zamani, and Jiawei Han. 2025. [Search-r1: Training llms to reason and leverage search engines with reinforcement learning](https://md.research.elamur.ai/free/en/papers/jin-2025-search-r1-training-llms-reason/). arXiv preprint arXiv:2503.09516.
14. Yujia Qin, Shengding Hu, Yankai Lin, Weize Chen, Ning Ding, Ganqu Cui, Zheni Zeng, Xuanhe Zhou, Yufei Huang, Chaojun Xiao, and 1 others. 2024. [Tool learning with foundation models](https://md.research.elamur.ai/free/en/papers/qin-2023-tool-learning-foundation-models/). ACM Computing Surveys, 57(4):1–40.
15. Yixin Li and 1 others. 2023. Gte-large model card. [https://huggingface.co/thenlper/gte-large](https://huggingface.co/thenlper/gte-large). Accessed: 2025-05-18. [↗](https://huggingface.co/thenlper/gte-large)
16. Yujia Qin, Shihao Liang, Yining Ye, Kunlun Zhu, Lan Yan, Yaxi Lu, Yankai Lin, Xin Cong, Xiangru Tang, Bill Qian, and 1 others. 2023. [Toolllm: Facilitating large language models to master 16000+ real-world apis](https://md.research.elamur.ai/free/en/papers/qin-2023-toolllm-facilitating-large-language-models/). arXiv preprint arXiv:2307.16789.
17. Weiwen Liu, Xu Huang, Xingshan Zeng, Xinlong Hao, Shuai Yu, Dexun Li, Shuai Wang, Weinan Gan, Zhengying Liu, Yuanqing Yu, and 1 others. 2024a. Toolace: Winning the points of llm function calling. arXiv preprint arXiv:2409.00920. [find](https://scholar.google.com/scholar?q=Weiwen+Liu%2C+Xu+Huang%2C+Xingshan+Zeng%2C+Xinlong+Hao%2C+Shuai+Yu%2C+Dexun+Li%2C+Shuai+Wang%2C+Weinan+Gan%2C+Zhengying+Liu%2C+Yuanqing+Yu%2C+and+1+others%2E+2024a%2E+Toolace%3A+Winning+the+points+of+llm+function+calling%2E+arXi)
18. Xukun Liu, Zhiyuan Peng, Xiaoyuan Yi, Xing Xie, Lirong Xiang, Yuchen Liu, and Dongkuan Xu. 2024b. Toolnet: Connecting large language models with massive tools via tool graph. arXiv preprint arXiv:2403.00839. — [arXiv:2403.00839](https://arxiv.org/abs/2403.00839) [find](https://scholar.google.com/scholar?q=Xukun+Liu%2C+Zhiyuan+Peng%2C+Xiaoyuan+Yi%2C+Xing+Xie%2C+Lirong+Xiang%2C+Yuchen+Liu%2C+and+Dongkuan+Xu%2E+2024b%2E+Toolnet%3A+Connecting+large+language+models+with+massive+tools+via+tool+graph%2E+arXiv+preprint+arXiv%3A2403)
19. Elias Lumer, Anmol Gulati, Vamse Kumar Subbiah, Pradeep Honaganahalli Basavaraju, and James A Burke. 2025. Scalemcp: Dynamic and auto-synchronizing model context protocol tools for llm agents. arXiv preprint arXiv:2505.06416. — [arXiv:2505.06416](https://arxiv.org/abs/2505.06416) [find](https://scholar.google.com/scholar?q=Elias+Lumer%2C+Anmol+Gulati%2C+Vamse+Kumar+Subbiah%2C+Pradeep+Honaganahalli+Basavaraju%2C+and+James+A+Burke%2E+2025%2E+Scalemcp%3A+Dynamic+and+auto%2Dsynchronizing+model+context+protocol+tools+for+llm+agents%2E+arXiv+p)
20. Elias Lumer, Vamse Kumar Subbiah, James A Burke, Pradeep Honaganahalli Basavaraju, and Austin Huber. 2024. Toolshed: Scale tool-equipped agents with advanced rag-tool fusion and tool knowledge bases. arXiv preprint arXiv:2410.14594. — [arXiv:2410.14594](https://arxiv.org/abs/2410.14594) [find](https://scholar.google.com/scholar?q=Elias+Lumer%2C+Vamse+Kumar+Subbiah%2C+James+A+Burke%2C+Pradeep+Honaganahalli+Basavaraju%2C+and+Austin+Huber%2E+2024%2E+Toolshed%3A+Scale+tool%2Dequipped+agents+with+advanced+rag%2Dtool+fusion+and+tool+knowledge+bases%2E+)
21. Meta Platforms, Inc. 2024. Model cards and prompt formats – llama 3.3. Accessed: 2025-05-18. [find](https://scholar.google.com/scholar?q=Meta+Platforms%2C+Inc%2E+2024%2E+Model+cards+and+prompt+formats+%E2%80%93+llama+3%2E3%2E+Accessed%3A+2025%2D05%2D18%2E)
22. Reiichiro Nakano, Jacob Hilton, Suchir Balaji, Jeff Wu, Long Ouyang, Christina Kim, Christopher Hesse, Shantanu Jain, Vineet Kosaraju, William Saunders, and 1 others. 2021. [Webgpt: Browser-assisted question-answering with human feedback](https://md.research.elamur.ai/free/en/papers/nakano-2021-webgpt-browser-assisted-question-answering/). arXiv preprint arXiv:2112.09332.
23. Nexusflow. 2023. Nexusraven-v2: Surpassing gpt-4 for zero-shot function calling. [https://nexusflow.ai/blogs/ravenv2](https://nexusflow.ai/blogs/ravenv2). Accessed: 2025-05-04. [↗](https://nexusflow.ai/blogs/ravenv2)
24. OpenAI. 2024. Hello gpt-4o. Accessed: 2025-05-18. [find](https://scholar.google.com/scholar?q=OpenAI%2E+2024%2E+Hello+gpt%2D4o%2E+Accessed%3A+2025%2D05%2D18%2E)
25. OpenAI. 2024. A practical guide to building agents. [https://cdn.openai.com/business-guides-and-resources/](https://cdn.openai.com/business-guides-and-resources/) a-practical-guide-to-building-agents.pdf. Accessed: 2025-05-16. [↗](https://cdn.openai.com/business-guides-and-resources/)
26. Shishir G Patil, Tianjun Zhang, Xin Wang, and Joseph E Gonzalez. 2024. [Gorilla: Large language model connected with massive apis](https://md.research.elamur.ai/free/en/papers/patil-2023-gorilla-large-language-model-connected/). Advances in Neural Information Processing Systems, 37:126544–126565.
27. Changle Qu, Sunhao Dai, Xiaochi Wei, Hengyi Cai, Shuaiqiang Wang, Dawei Yin, Jun Xu, and Ji-Rong Wen. 2024. From exploration to mastery: enabling llms to master tools via self-driven interactions. arXiv preprint arXiv:2410.08197. — [arXiv:2410.08197](https://arxiv.org/abs/2410.08197) [find](https://scholar.google.com/scholar?q=Changle+Qu%2C+Sunhao+Dai%2C+Xiaochi+Wei%2C+Hengyi+Cai%2C+Shuaiqiang+Wang%2C+Dawei+Yin%2C+Jun+Xu%2C+and+Ji%2DRong+Wen%2E+2024%2E+From+exploration+to+mastery%3A+enabling+llms+to+master+tools+via+self%2Ddriven+interactions%2E+arX)
28. Changle Qu, Sunhao Dai, Xiaochi Wei, Hengyi Cai, Shuaiqiang Wang, Dawei Yin, Jun Xu, and Ji-Rong Wen. 2025. Tool learning with [large language models: A survey](https://md.research.elamur.ai/free/en/papers/minaee-2024-large-language-models-survey/). Frontiers of Computer Science, 19(8):198343.
29. Stephen Robertson, Hugo Zaragoza, and 1 others. 2009. The probabilistic relevance framework: Bm25 and beyond. Foundations and Trends® in Information Retrieval, 3(4):333–389. [find](https://scholar.google.com/scholar?q=Stephen+Robertson%2C+Hugo+Zaragoza%2C+and+1+others%2E+2009%2E+The+probabilistic+relevance+framework%3A+Bm25+and+beyond%2E+Foundations+and+Trends%C2%AE+in+Information+Retrieval%2C+3%284%29%3A333%E2%80%93389%2E)
30. Peter J Rousseeuw. 1987. Silhouettes: a graphical aid to the interpretation and validation of cluster analysis. Journal of computational and applied mathematics, 20:53–65. [find](https://scholar.google.com/scholar?q=Peter+J+Rousseeuw%2E+1987%2E+Silhouettes%3A+a+graphical+aid+to+the+interpretation+and+validation+of+cluster+analysis%2E+Journal+of+computational+and+applied+mathematics%2C+20%3A53%E2%80%9365%2E)
31. Yongliang Shen, Kaitao Song, Xu Tan, Wenqi Zhang, Kan Ren, Siyu Yuan, Weiming Lu, Dongsheng Li, and Yueting Zhuang. 2024. Taskbench: Benchmarking large language models for task automation. Advances in Neural Information Processing Systems, 37:4540–4574. [find](https://scholar.google.com/scholar?q=Yongliang+Shen%2C+Kaitao+Song%2C+Xu+Tan%2C+Wenqi+Zhang%2C+Kan+Ren%2C+Siyu+Yuan%2C+Weiming+Lu%2C+Dongsheng+Li%2C+and+Yueting+Zhuang%2E+2024%2E+Taskbench%3A+Benchmarking+large+language+models+for+task+automation%2E+Advances+in)
32. Zhuocheng Shen. 2024. Llm with tools: A survey. arXiv preprint arXiv:2409.18807. — [arXiv:2409.18807](https://arxiv.org/abs/2409.18807) [find](https://scholar.google.com/scholar?q=Zhuocheng+Shen%2E+2024%2E+Llm+with+tools%3A+A+survey%2E+arXiv+preprint+arXiv%3A2409%2E18807%2E+%E2%80%94+%5BarXiv%3A2409%2E18807%5D%28https%3A%2F%2Farxiv%2Eorg%2Fabs%2F2409%2E18807%29)
33. Yifan Song, Weimin Xiong, Dawei Zhu, Wenhao Wu, Han Qian, Mingbo Song, Hailiang Huang, Cheng Li, Ke Wang, Rong Yao, and 1 others. 2023. [Restgpt: Connecting large language models with real-world restful apis](https://md.research.elamur.ai/free/en/papers/song-2023-restgpt-connecting-large-language-models/). arXiv preprint arXiv:2306.06624.
34. Qiaoyu Tang, Ziliang Deng, Hongyu Lin, Xianpei Han, Qiao Liang, Boxi Cao, and Le Sun. 2023. [Toolalpaca: Generalized tool learning for language models with 3000 simulated cases](https://md.research.elamur.ai/free/en/papers/tang-2023-toolalpaca-generalized-tool-learning-language/). arXiv preprint arXiv:2306.05301.
35. Laurens Van der Maaten and Geoffrey Hinton. 2008. Visualizing data using t-sne. Journal of machine learning research, 9(11). 34104 10 [find](https://scholar.google.com/scholar?q=Laurens+Van+der+Maaten+and+Geoffrey+Hinton%2E+2008%2E+Visualizing+data+using+t%2Dsne%2E+Journal+of+machine+learning+research%2C+9%2811%29%2E+34104+10)
36. Hongru Wang, Cheng Qian, Wanjun Zhong, Xiusi Chen, Jiahao Qiu, Shijue Huang, Bowen Jin, Mengdi Wang, Kam-Fai Wong, and Heng Ji. 2025. Otc: Optimal tool calls via reinforcement learning. arXiv preprint arXiv:2504.14870. — [arXiv:2504.14870](https://arxiv.org/abs/2504.14870) [find](https://scholar.google.com/scholar?q=Hongru+Wang%2C+Cheng+Qian%2C+Wanjun+Zhong%2C+Xiusi+Chen%2C+Jiahao+Qiu%2C+Shijue+Huang%2C+Bowen+Jin%2C+Mengdi+Wang%2C+Kam%2DFai+Wong%2C+and+Heng+Ji%2E+2025%2E+Otc%3A+Optimal+tool+calls+via+reinforcement+learning%2E+arXiv+preprint)
37. Cailin Winston and René Just. A taxonomy of failures in tool-augmented llms. [find](https://scholar.google.com/scholar?q=Cailin+Winston+and+Ren%C3%A9+Just%2E+A+taxonomy+of+failures+in+tool%2Daugmented+llms%2E)
38. Mengsong Wu, Tong Zhu, Han Han, Chuanyuan Tan, Xiang Zhang, and Wenliang Chen. 2024a. Seal-tools: Self-instruct tool learning dataset for agent tuning and detailed benchmark. In CCF International Conference on Natural Language Processing and Chinese Computing, pages 372–384. Springer. [find](https://scholar.google.com/scholar?q=Mengsong+Wu%2C+Tong+Zhu%2C+Han+Han%2C+Chuanyuan+Tan%2C+Xiang+Zhang%2C+and+Wenliang+Chen%2E+2024a%2E+Seal%2Dtools%3A+Self%2Dinstruct+tool+learning+dataset+for+agent+tuning+and+detailed+benchmark%2E+In+CCF+International+Conf)
39. Shirley Wu, Shiyu Zhao, Qian Huang, Kexin Huang, Michihiro Yasunaga, Kaidi Cao, Vassilis Ioannidis, Karthik Subbian, Jure Leskovec, and James Y Zou. 2024b. Avatar: Optimizing llm agents for tool usage via contrastive reasoning. Advances in Neural Information Processing Systems, 37:25981–26010. [find](https://scholar.google.com/scholar?q=Shirley+Wu%2C+Shiyu+Zhao%2C+Qian+Huang%2C+Kexin+Huang%2C+Michihiro+Yasunaga%2C+Kaidi+Cao%2C+Vassilis+Ioannidis%2C+Karthik+Subbian%2C+Jure+Leskovec%2C+and+James+Y+Zou%2E+2024b%2E+Avatar%3A+Optimizing+llm+agents+for+tool+usage)
40. Qiantong Xu, Fenglu Hong, Bo Li, Changran Hu, Zhengyu Chen, and Jian Zhang. 2023. [On the tool manipulation capability of open-source large language models](https://md.research.elamur.ai/free/en/papers/xu-2023-tool-manipulation-capability-open-source/). arXiv preprint arXiv:2305.16504.
41. Rui Yang, Lin Song, Yanwei Li, Sijie Zhao, Yixiao Ge, Xiu Li, and Ying Shan. 2023. Gpt4tools: Teaching large language model to use tools via self-instruction. Advances in Neural Information Processing Systems, 36:71995–72007. [find](https://scholar.google.com/scholar?q=Rui+Yang%2C+Lin+Song%2C+Yanwei+Li%2C+Sijie+Zhao%2C+Yixiao+Ge%2C+Xiu+Li%2C+and+Ying+Shan%2E+2023%2E+Gpt4tools%3A+Teaching+large+language+model+to+use+tools+via+self%2Dinstruction%2E+Advances+in+Neural+Information+Processing)
42. Asaf Yehudai, Lilach Eden, Alan Li, Guy Uziel, Yilun Zhao, Roy Bar-Haim, Arman Cohan, and Michal Shmueli-Scheuer. 2025. Survey on evaluation of llm-based agents. arXiv preprint arXiv:2503.16416. — [arXiv:2503.16416](https://arxiv.org/abs/2503.16416) [find](https://scholar.google.com/scholar?q=Asaf+Yehudai%2C+Lilach+Eden%2C+Alan+Li%2C+Guy+Uziel%2C+Yilun+Zhao%2C+Roy+Bar%2DHaim%2C+Arman+Cohan%2C+and+Michal+Shmueli%2DScheuer%2E+2025%2E+Survey+on+evaluation+of+llm%2Dbased+agents%2E+arXiv+preprint+arXiv%3A2503%2E16416%2E+%E2%80%94+%5Bar)
43. Lifan Yuan, Yangyi Chen, Xingyao Wang, Yi R Fung, Hao Peng, and Heng Ji. 2023. Craft: Customizing llms by creating and retrieving from specialized toolsets. arXiv preprint arXiv:2309.17428. — [arXiv:2309.17428](https://arxiv.org/abs/2309.17428) [find](https://scholar.google.com/scholar?q=Lifan+Yuan%2C+Yangyi+Chen%2C+Xingyao+Wang%2C+Yi+R+Fung%2C+Hao+Peng%2C+and+Heng+Ji%2E+2023%2E+Craft%3A+Customizing+llms+by+creating+and+retrieving+from+specialized+toolsets%2E+arXiv+preprint+arXiv%3A2309%2E17428%2E+%E2%80%94+%5BarXiv%3A2)

*Updated 31 October 2026*