# TTPA: Token-level Tool-use Preference Alignment Training Framework with Fine-grained Evaluation

arXiv: https://arxiv.org/abs/2505.20016 · HTML: https://mlsanity.com/paper/2505.20016/

## Key facts

- Cited by 9 papers in the map (data through 14 Oct 2026).
- Placed in Tool learning and function calling, one of 16 directions in LLM agents and tool use.
- Builds on 28 of its 38 references that are in the map.
- First posted on arXiv 26 May 2025.

## Placed in

- LLM agents and tool use / [Tool learning and function calling](https://mlsanity.com/agents/tool-learning-and-function-calling/)
- LLM alignment / RLHF / preference optimisation / [LLM preference optimization methods](https://mlsanity.com/alignment/llm-preference-optimization-methods/)

## Builds on (most cited first, 28 of 28)

- [Proximal Policy Optimization Algorithms](https://mlsanity.com/paper/1707.06347/) (2017)
- [Llama 2: Open Foundation and Fine-Tuned Chat Models](https://mlsanity.com/paper/2307.09288/) (2023)
- [Training Verifiers to Solve Math Word Problems](https://mlsanity.com/paper/2110.14168/) (2021)
- [ReAct: Synergizing Reasoning and Acting in Language Models](https://mlsanity.com/paper/2210.03629/) (2022)
- [Direct Preference Optimization: Your Language Model is Secretly a Reward Model](https://mlsanity.com/paper/2305.18290/) (2023)
- [Toolformer: Language Models Can Teach Themselves to Use Tools](https://mlsanity.com/paper/2302.04761/) (2023)
- [HellaSwag: Can a Machine Really Finish Your Sentence?](https://mlsanity.com/paper/1905.07830/) (2019)
- [ToolLLM: Facilitating Large Language Models to Master 16000+ Real-world APIs](https://mlsanity.com/paper/2307.16789/) (2023)
- [MMLU-Pro: A More Robust and Challenging Multi-Task Language Understanding Benchmark](https://mlsanity.com/paper/2406.01574/) (2024)
- [CommonsenseQA: A Question Answering Challenge Targeting Commonsense Knowledge](https://mlsanity.com/paper/1811.00937/) (2018)
- [Gorilla: Large Language Model Connected with Massive APIs](https://mlsanity.com/paper/2305.15334/) (2023)
- [Plug and Play Language Models: A Simple Approach to Controlled Text Generation](https://mlsanity.com/paper/1912.02164/) (2019)
- [ToolAlpaca: Generalized Tool Learning for Language Models with 3000 Simulated Cases](https://mlsanity.com/paper/2306.05301/) (2023)
- [TravelPlanner: A Benchmark for Real-World Planning with Language Agents](https://mlsanity.com/paper/2402.01622/) (2024)
- [MINT: Evaluating LLMs in Multi-turn Interaction with Tools and Language Feedback](https://mlsanity.com/paper/2309.10691/) (2023)
- [MetaTool Benchmark for Large Language Models: Deciding Whether to Use Tools and Which to Use](https://mlsanity.com/paper/2310.03128/) (2023)
- [ToolACE: Winning the Points of LLM Function Calling](https://mlsanity.com/paper/2409.00920/) (2024)
- [APIGen: Automated Pipeline for Generating Verifiable and Diverse Function-Calling Datasets](https://mlsanity.com/paper/2406.18518/) (2024)
- [Opportunities and Challenges for ChatGPT and Large Language Models in Biomedicine and Health](https://mlsanity.com/paper/2306.10070/) (2023)
- [A Thorough Examination of Decoding Methods in the Era of LLMs](https://mlsanity.com/paper/2402.06925/) (2024)
- [Confucius: Iterative Tool Learning from Introspection Feedback by Easy-to-Difficult Curriculum](https://mlsanity.com/paper/2308.14034/) (2023)
- [Hammer: Robust Function-Calling for On-Device Language Models via Function Masking](https://mlsanity.com/paper/2410.04587/) (2024)
- [Advancing Tool-Augmented Large Language Models: Integrating Insights from Errors in Inference Trees](https://mlsanity.com/paper/2406.07115/) (2024)
- [ToolPlanner: A Tool Augmented LLM for Multi Granularity Instructions with Path Planning and Feedback](https://mlsanity.com/paper/2409.14826/) (2024)
- [TL-Training: A Task-Feature-Based Framework for Training Large Language Models in Tool Use](https://mlsanity.com/paper/2412.15495/) (2024)
- [ToolComp: A Multi-Tool Reasoning & Process Supervision Benchmark](https://mlsanity.com/paper/2501.01290/) (2025)
- [ADC: Enhancing Function Calling Via Adversarial Datasets and Code Line-Level Feedback](https://mlsanity.com/paper/2412.17754/) (2024)
- [What Affects the Stability of Tool Learning? An Empirical Study on the Robustness of Tool Learning Frameworks](https://mlsanity.com/paper/2407.03007/) (2024)

## Built on by (most cited first, 9 of 9)

- [A Survey of Context Engineering for Large Language Models](https://mlsanity.com/paper/2507.13334/) (2025)
- [In-the-Flow Agentic System Optimization for Effective Planning and Tool Use](https://mlsanity.com/paper/2510.05592/) (2025)
- [Are Tools All We Need? Unveiling the Tool-Use Tax in LLM Agents](https://mlsanity.com/paper/2605.00136/) (2026)
- [FunReason-MT Technical Report: Advanced Data Synthesis Solution for Real-world Multi-Turn Tool-use](https://mlsanity.com/paper/2510.24645/) (2025)
- [From Failure to Mastery: Generating Hard Samples for Tool-use Agents](https://mlsanity.com/paper/2601.01498/) (2026)
- [PACE: Prefix-Protected and Difficulty-Aware Compression for Efficient Reasoning](https://mlsanity.com/paper/2602.11639/) (2026)
- [DPEPO: Diverse Parallel Exploration Policy Optimization for LLM-based Agents](https://mlsanity.com/paper/2604.24320/) (2026)
- [HTAA: Enhancing LLM Planning via Hybrid Toolset Agentization & Adaptation](https://mlsanity.com/paper/2604.10917/) (2026)
- [Agentic Multi-Turn Reasoning: A Fairness Approach](https://mlsanity.com/paper/2609.33323/) (2026)

## Abstract (arXiv)

Existing tool-learning methods usually rely on supervised fine-tuning, they often overlook fine-grained optimization of internal tool call details, leading to limitations in preference alignment and error discrimination. To overcome these challenges, we propose Token-level Tool-use Preference Alignment Training Framework (TTPA), a training paradigm for constructing token-level tool-use preference datasets that align LLMs with fine-grained preferences using a novel error-oriented scoring mechanism. TTPA first introduces reversed dataset construction, a method for creating high-quality, multi-turn tool-use datasets by reversing the generation flow. Additionally, we propose Token-level Preference Sampling (TPS) to capture fine-grained preferences by modeling token-level differences during generation. To address biases in scoring, we introduce the Error-oriented Scoring Mechanism (ESM), which quantifies tool-call errors and can be used as a training signal. Extensive experiments on three diverse benchmark datasets demonstrate that TTPA significantly improves tool-using performance while showing strong generalization ability across models and datasets.