Research map / Large language models / Instruction and fine-tuning
Mining Useful General Data for Low-Resource Domain Adaptation
Pingjie Wang, Hongcheng Liu, Yusheng Liao, Ziqing Fan, Yaxin Du, Shuo Tang, Yanfeng Wang, Yu Wang
Read the paperCheck your idea against 137,217 papers
AI-extracted. The measured effects below were read by AI from the paper's full text (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.
Measured effects reported in the paper
| What was changed | Where | Measured | Effect |
|---|---|---|---|
| fine-tune on only 1K domain data | Llama3-8B-Instruct · ContractNLI | accuracy: 59.9 -> 41.1 | Hurt |
| fine-tune on only 1K domain data | Qwen3-8B · FPB | accuracy: 80.1 -> 72.2 | Hurt |
| add 9K randomly selected auxiliary data | Llama3-8B-Instruct · average of 8 tasks | accuracy: 68.7 -> 74.3 | Improved |
| add 9K randomly selected auxiliary data | Qwen3-8B · average of 8 tasks | accuracy: 74.4 -> 72.9 | Hurt |
| select auxiliary data with NTK-Selector instead of random | Llama3-8B-Instruct · average of 8 tasks | accuracy: 74.3 -> 76.6 | Improved |
| select auxiliary data with NTK-Selector instead of DSIR | Qwen3-8B · average of 8 tasks | accuracy: 77.7 -> 78.6 | Improved |
| select auxiliary data with LESS | Qwen3-8B · FPB | accuracy: 80.1 -> 60.3 | Hurt |
| increase total data from x1 to x5 with NTK-Selector auxiliary data (ablation) | NTK-Selector · TFNS | accuracy: 69.3 -> 71.9 | Improved |
| increase total data from x10 to x20 (ablation) | NTK-Selector · TFNS | accuracy: 73.3 -> 73.3 | No clear effect |
| increase pre-selection size from 2N to 16N (ablation) | NTK-Selector · MMLU-Med | accuracy: 73.3 -> 74.4 | Improved |
| increase pre-selection size from 2N to 16N (ablation) | NTK-Selector · FPB | accuracy: 84.5 -> 86.9 | Improved |
| increase pre-selection size from 4N to 16N (ablation) | NTK-Selector · RSDD | accuracy: 96.5 -> 95.9 | No clear effect |
Closest papers by meaning
- Boosting Multi-Domain Fine-Tuning of Large Language Models through Evolving Interactions between Samples (ICML 2025 · similarity 0.89)
- Auto-GDA: Automatic Domain Adaptation for Efficient Grounding Verification in Retrieval-Augmented Generation (ICLR 2025 · similarity 0.88)
- BLADE: Enhancing Black-Box Large Language Models with Small Domain-Specific Models (AAAI 2025 · similarity 0.88)
- Data Efficient Adaptation in Large Language Models via Continuous Low-Rank Fine-Tuning (NeurIPS 2025 · similarity 0.88)
- Language-Induced Priors for Domain Adaptation (NeurIPS 2026 · similarity 0.88)
- Get more for less: Principled Data Selection for Warming Up Fine-Tuning in LLMs (ICLR 2024 · similarity 0.88)
Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.
