Lune

ACL Findings2025

APT: Improving Specialist LLM Performance with Weakness Case Acquisition and Iterative Preference Training

Jun Rao, Zepeng Lin, Xuebo Liu, Xiaopeng Ke, Lian Lian, Dong Jin, Shengjun Cheng, Jun Yu, Min Zhang

Abstract

Large Language Models (LLMs) often require domain-specific fine-tuning to address targeted tasks, which risks degrading their general capabilities. Maintaining a balance between domain-specific enhancements and general model utility is a key challenge. This paper proposes a novel approach named APT (Weakness Case Acquisition and Iterative Preference Training) to enhance domain-specific performance with self-generated dis-preferred weakness data (bad cases and similar cases). APT uniquely focuses on training the model using only those samples where errors occur, alongside a small, similar set of samples retrieved for this purpose. This targeted training minimizes interference with the model's existing knowledge base, effectively retaining generic capabilities. Experimental results on the LLama-2 and Mistral-V0.3 models across various benchmarks demonstrate that APT ensures no reduction in generic capacity and achieves superior performance on downstream tasks compared to various existing methods. This validates our method as an effective strategy for enhancing domain-specific capabilities without sacrificing the model's broader applicability.

Ask Lune

Know this paper in minutes, not an afternoon.

Lune can read every page of it and the top-tier papers that cite it, and gives you its claims, its evidence and its limits, each one cited to the paper it comes from. No skimming, no AI slop.

6 top-tier papers have cited it since. Know what they found before your reviewers do.

Questions to start from

Lune can draw on

the full paper · the top-tier papers that cite it · related work

Get the answer

Free to start. No credit card required.

Lineage

  • ProbeLLM: Automating Principled Diagnosis of LLM Failures cites Measuring Massive Multitask Language Understanding
  • Merge then Realign: Simple and Effective Modality-Incremental Continual Learning for Multimodal LLMs cites LoRA: Low-Rank Adaptation of Large Language Models
Builds onCited byLeft to right by year; the more cited, the higher.

Cited by top-tier papers6

  1. KaLM-Embedding-V2: Superior Training Techniques and Data Inspire A Versatile Embedding ModelXinping Zhao, Xinshuo Hu, Zifei Shan, Shouzheng Huang et al.ICLR 2026 · 47 citations
  2. ProbeLLM: Automating Principled Diagnosis of LLM FailuresYue Huang, Zhengzhe Jiang, Yuchen Ma, Yu Jiang et al.ICML 2026 · 4 citations
  3. Merge then Realign: Simple and Effective Modality-Incremental Continual Learning for Multimodal LLMsDingkun Zhang, Shuhan Qi, Xinyu Xiao, Kehai Chen et al.EMNLP 2025 · 4 citations
  4. DeReA: Improving Idiom Translation with Detect-Retrieve-Arbitrate ReasoningRongqing Jiang, Xuebo Liu, Shengxin Liu, Yutong Wang et al.ACL 2026
  5. SEAL: Structure and Element Aware Learning Improves Long Structured Document RetrievalXinhao Huang, Zhibo Ren, Yipeng Yu, Ying Zhou et al.EMNLP 2025

Builds on31

  1. Chain-of-Thought Prompting Elicits Reasoning in Large Language ModelsJason Wei, Xuezhi Wang, Dale Schuurmans, Maarten Bosma et al.NeurIPS 2022 · 22,562 citations
  2. LoRA: Low-Rank Adaptation of Large Language ModelsEdward J. Hu, Yelong Shen, Phillip Wallis, Zeyuan Allen-Zhu et al.ICLR 2022 · 18,833 citations
  3. Direct Preference Optimization: Your Language Model is Secretly a Reward ModelRafael Rafailov, Archit Sharma, Eric Mitchell, Christopher D. Manning et al.NeurIPS 2023 · 10,924 citations
  4. Measuring Massive Multitask Language UnderstandingDan Hendrycks, Collin Burns, Steven Basart, Andy Zou et al.ICLR 2021 · 7,905 citations
  5. WinoGrande: An Adversarial Winograd Schema Challenge at ScaleKeisuke Sakaguchi, Ronan Le Bras, Chandra Bhagavatula, Yejin ChoiAAAI 2020 · 3,037 citations

Related papers

  1. PANDA: Preference Adaptation for Enhancing Domain-Specific Abilities of LLMsAn Liu, Zonghan Yang, Zhenhe Zhang, Qingyuan Hu et al.ACL Findings 2024
  2. Demystifying Domain-adaptive Post-training for Financial LLMsZixuan Ke, Yifei Ming, Xuan-Phi Nguyen, Caiming Xiong et al.EMNLP 2025 · 2 citations
  3. Task Oriented In-Domain Data AugmentationXiao Liang, Xinyu Hu, Simiao Zuo, Yeyun Gong et al.EMNLP 2024 · 1 citation
  4. LLM2LLM: Boosting LLMs with Novel Iterative Data EnhancementNicholas Lee, Thanakul Wattanawong, Sehoon Kim, Karttikeya Mangalam et al.ACL Findings 2024 · 21 citations
  5. Improving Influence-based Instruction Tuning Data Selection for Balanced Learning of Diverse CapabilitiesQirun Dai, Dylan Zhang, Jiaqi W. Ma, Hao PengEMNLP Findings 2025 · 1 citation
Dusk over the sea between two cliffs drawn in fine vertical lines