The reviewed record of science sign in
Pith

arxiv: 2402.02318 · v1 · pith:Q4NY3TDS · submitted 2024-02-04 · cs.LG · cs.CL

Diversity Measurement and Subset Selection for Instruction Tuning Datasets

Peiqi Wang , Yikang Shen , Zhen Guo , Matthew Stallone , Yoon Kim , Polina Golland , Rameswar Panda This is my paper

Reviewed by Pithpith:Q4NY3TDSopen to challenge →

classification cs.LG cs.CL
keywords datasetdiversitydatasetsinstructionselectiontuningcurationdata
0
0 comments X
read the original abstract

We aim to select data subsets for the fine-tuning of large language models to more effectively follow instructions. Prior work has emphasized the importance of diversity in dataset curation but relied on heuristics such as the number of tasks. In this paper, we use determinantal point processes to capture the diversity and quality of instruction tuning datasets for subset selection. We propose to measure dataset diversity with log determinant distance that is the distance between the dataset of interest and a maximally diverse reference dataset. Our experiments demonstrate that the proposed diversity measure in the normalized weight gradient space is correlated with downstream instruction-following performance. Consequently, it can be used to inform when data selection is the most helpful and to analyze dataset curation strategies. We demonstrate the utility of our approach on various instruction tuning datasets.

This paper has not been read by Pith yet.

discussion (0)

Sign in with ORCID, Apple, or X to comment. Anyone can read and Pith papers without signing in.

Forward citations

Cited by 4 Pith papers

Reviewed papers in the Pith corpus that reference this work. Sorted by Pith novelty score.

  1. InduceKV: Fixed-Footprint Continual Adaptation of Multimodal LLMs via Inducing KV Memories

    cs.AI 2026-07 unverdicted novelty 6.0

    InduceKV is a retrieval-based continual adaptation method that uses bilevel selection to build a compact set of inducing KV memories for fixed-footprint updates to multimodal LLMs.

  2. What Makes Good Instruction-Tuning Data? An In-Context Learning Perspective

    cs.CL 2026-04 unverdicted novelty 6.0

    A weighted in-context influence metric selects effective instruction-tuning data, outperforming baselines while showing that harder samples have lower influence.

  3. Cram Less to Fit More: Training Data Pruning Improves Memorization of Facts

    cs.CL 2026-04 conditional novelty 6.0

    Loss-based pruning of training data to limit facts and flatten their frequency distribution enables a 110M-parameter GPT-2 model to memorize 1.3 times more entity facts than standard training, matching a 1.3B-paramete...

  4. DUET: Optimizing Training Data Mixtures via Feedback from Unseen Evaluation Tasks

    cs.LG 2025-02 unverdicted novelty 6.0

    DUET is a global-to-local method that optimizes LLM training data mixtures via Bayesian optimization guided by influence-based selection and feedback from unseen evaluation tasks, with a regret bound showing convergen...