Lune

ACL Findings2024

RECOST: External Knowledge Guided Data-efficient Instruction Tuning

Qi Zhang, Yiming Zhang, Haobo Wang, Junbo Zhao

Abstract

In the current landscape of large language models (LLMs), the process of instruction tuning serves as an essential step. Considering the high computing power overhead, dataefficient instruction tuning was proposed to reduce the training data size in this process, aiming at selecting high-quality instructional data. Nevertheless, we argue that most current data-efficient instruction-tuning methods are highly dependent on the quality of the original instruction-tuning dataset. When it comes to datasets synthesized by LLMs, a common scenario in this field, dirty samples will even be selected with a higher probability than other samples. To address these challenges, we utilized external knowledge (relevant examples or paragraphs) to evaluate those samples synthesized by LLMs with an in-context-based relative predictive entropy. Based on the new metric, we proposed a framework, dubbed as RECOST, which integrates external-knowledge-base reranking and diversity-consistent sampling into a single pipeline. Through extensive experiments on several synthetic datasets (Alpaca and Alpaca-gpt4), we demonstrate the effectiveness of our method and achieve even better results with only 1% of the full dataset.

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.

4 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

  • Towards Acyclic Preference Evaluation of Language Models via Multiple Evaluators cites Training language models to follow instructions with human feedback
  • Measuring Massive Multitask Language Understanding cites Language Models are Few-Shot Learners
  • TRACE: Trajectory-based Activation Change Estimation for Task-specific Data Selection cites Measuring Massive Multitask Language Understanding
  • TruthfulQA: Measuring How Models Mimic Human Falsehoods cites Measuring Massive Multitask Language Understanding
  • TruthfulQA: Measuring How Models Mimic Human Falsehoods cites Training language models to follow instructions with human feedback
  • Training language models to follow instructions with human feedback cites TruthfulQA: Measuring How Models Mimic Human Falsehoods
  • Automatically Generating Numerous Context-Driven SFT Data for LLMs Across Diverse Granularity cites Training language models to follow instructions with human feedback
  • Finetuned Language Models are Zero-Shot Learners cites Language Models are Few-Shot Learners
  • Finetuned Language Models are Zero-Shot Learners cites Training language models to follow instructions with human feedback
  • Training language models to follow instructions with human feedback cites Finetuned Language Models are Zero-Shot Learners
  • I3: Intent-Introspective Retrieval Conditioned on Instructions cites Training language models to follow instructions with human feedback
  • Training language models to follow instructions with human feedback cites Language Models are Few-Shot Learners
Builds onCited byLeft to right by year; the more cited, the higher.

Cited by top-tier papers4

  1. Towards Acyclic Preference Evaluation of Language Models via Multiple EvaluatorsZhengyu Hu, Jieyu Zhang, Zhihan Xiong, Alexander Ratner et al.AAAI 2026 · 14 citations
  2. Automatically Generating Numerous Context-Driven SFT Data for LLMs Across Diverse GranularityShanghaoran QuanAAAI 2025 · 7 citations
  3. I3: Intent-Introspective Retrieval Conditioned on InstructionsKaihang Pan, Juncheng Li, Wenjie Wang, Hao Fei et al.SIGIR 2024 · 7 citations
  4. TRACE: Trajectory-based Activation Change Estimation for Task-specific Data SelectionYe He, Shangzhan Li, Yuxin Zhou, Qi ShiAAAI 2026

Builds on11

  1. Language Models are Few-Shot LearnersTom B. Brown, Benjamin Mann, Nick Ryder, Melanie Subbiah et al.NeurIPS 2020 · 64,255 citations
  2. Training language models to follow instructions with human feedbackLong Ouyang, Jeffrey Wu, Xu Jiang, Diogo Almeida et al.NeurIPS 2022 · 24,707 citations
  3. Measuring Massive Multitask Language UnderstandingDan Hendrycks, Collin Burns, Steven Basart, Andy Zou et al.ICLR 2021 · 7,905 citations
  4. Finetuned Language Models are Zero-Shot LearnersJason Wei, Maarten Bosma, Vincent Y. Zhao, Kelvin Guu et al.ICLR 2022 · 4,966 citations
  5. TruthfulQA: Measuring How Models Mimic Human FalsehoodsStephanie Lin, Jacob Hilton, Owain EvansACL 2022 · 3,228 citations

Related papers

  1. From Selection to Refinement: Iterative Optimization for Instruction DataHang Hu, Ziyan Liu, Rujie Wen, Ruihui Hou et al.ACL 2026
  2. Clustering and Ranking: Diversity-preserved Instruction Selection through Expert-aligned Quality EstimationYuan Ge, Yilun Liu, Chi Hu, Weibin Meng et al.EMNLP 2024 · 10 citations
  3. What Makes Good Data for Alignment? A Comprehensive Study of Automatic Data Selection in Instruction TuningWei Liu, Weihao Zeng, Keqing He, Yong Jiang et al.ICLR 2024 · 369 citations
  4. BRIEF: Bi-level Coreset Selection for Efficient Instruction Tuning in LLMsChaoyuan Shen, Chi Zhang, Chengliang Chai, Jiacheng Wang et al.VLDB 2026 · 2 citations
  5. Not All Documents Are What You Need for Extracting Instruction Tuning DataChi Zhang, Huaping Zhong, Hongtao Li, Chengliang Chai et al.ICLR 2026 · 3 citations
Dusk over the sea between two cliffs drawn in fine vertical lines