DataPrep-Bench: Benchmarking LLMs as Training Data Preparators

• URL: http://arxiv.org/abs/2607.20465v1
• Date: Tue, 19 May 2026 02:23:53 GMT
• Title: DataPrep-Bench: Benchmarking LLMs as Training Data Preparators
• Abstract summary: We introduce DataPrep-Bench, the first unified benchmark that jointly evaluates data construction and data quality evaluation.<n>For data construction, methods consume identical raw sources and are scored by fine-tuning a base model on their outputs jointly with Dolly-15k.<n>For data quality evaluation, scoring functions are scored by Pearson correlation with downstream performance on a shared candidate pool.
• Score: 35.76000147244654
• License: http://creativecommons.org/licenses/by/4.0/
• Abstract: The quality of training data fundamentally determines the capabilities of large language models (LLMs), yet no unified benchmark exists to measure how well LLMs, agents, and data-centric workflows actually prepare training data end to end. We view LLM-driven data preparation as comprising two complementary capabilities: data construction, which transforms raw sources into supervised training data, and data quality evaluation, which predicts the training value of candidate datasets before downstream training; throughout, "quality" refers to downstream training utility rather than surface-level textual properties. We introduce DataPrep-Bench, the first unified benchmark that jointly evaluates both capabilities under a shared downstream-grounded protocol over six domains and multiple base models. For data construction, methods consume identical raw sources and are scored by fine-tuning a base model on their outputs jointly with Dolly-15k; alongside this track we release Data-Construction-Skill, a skill-guided agent that lifts the Dolly-only baseline by nearly 20 points absolute on Llama-3.1-8B Finance and is competitive with the strongest agent- and DataFlow-based methods in knowledge-extraction-dense domains. For data quality evaluation, scoring functions are scored by Pearson correlation with downstream performance on a shared candidate pool; we release the Distributional Alignment Score (DAS), a distribution-based evaluator that uses MMD between a candidate dataset and a domain proxy. DAS attains the strongest cross-model correlation in four of six domains and is the only metric clearing r > 0.70 simultaneously in Math, Science, and Medical, outperforming existing quality-, diversity-, and heuristic-based evaluators. DataPrep-Bench provides a unified, downstream-grounded framework for measuring progress on both capabilities as co-equal targets of LLM-driven data preparation.

Related papers

• DataMaster: Data-Centric Autonomous AI Research [51.95819363392062]
  We study task-conditioned autonomous data engineering, where an agent improves a fixed learning algorithm by optimizing only the data side.<n>We propose DataMaster, a data-agent framework that integrates tree-structured search, shared candidate data, and cumulative memory.<n>We evaluate DataMaster on two types of benchmarks, MLE-Bench Lite and PostTrainBench.
  arXiv  Detail & Related papers  (2026-05-11T17:46:24Z)
• OpenDataArena: A Fair and Open Arena for Benchmarking Post-Training Dataset Value [74.80873109856563]
  OpenDataArena (ODA) is a holistic and open platform designed to benchmark the intrinsic value of post-training data.<n>ODA establishes a comprehensive ecosystem comprising four key pillars: (i) a unified training-evaluation pipeline that ensures fair, open comparisons across diverse models; (ii) a multi-dimensional scoring framework that profiles data quality along tens of distinct axes; and (iii) an interactive data lineage explorer to visualize dataset genealogy and dissect component sources.
  arXiv  Detail & Related papers  (2025-12-16T03:33:24Z)
• Fixing It in Post: A Comparative Study of LLM Post-Training Data Quality and Model Performance [29.94723846950853]
  We conduct the first comprehensive analysis of two prominent open post-training datasets: Tulu-3-SFT-Mix and SmolTalk.<n>We derive statistics that reveal structural and qualitative similarities and differences between the two datasets.<n>Our findings offer actionable insights for constructing more effective post-training datasets.
  arXiv  Detail & Related papers  (2025-06-06T20:34:06Z)
• DataMan: Data Manager for Pre-training Large Language Models [39.677609311769146]
  Existing methods rely on limited intuition, lacking comprehensive and clear guidelines.<n>We derive 14 quality criteria from the causes of text perplexity anomalies and introduce 15 common application domains to support domain mixing.<n>Our experiments validate our approach, using DataMan to select 30B tokens to train a 1.3B- parameter language model.
  arXiv  Detail & Related papers  (2025-02-26T18:01:19Z)
• Improving Pretraining Data Using Perplexity Correlations [56.41097718862742]
  We present a framework that selects high-quality pretraining data without any LLM training of our own.<n>We build a new statistical framework for data selection centered around estimates of perplexity-benchmark correlations.<n>Our approach outperforms DSIR on every benchmark, while matching the best data selector found in DataComp-LM.
  arXiv  Detail & Related papers  (2024-09-09T17:23:29Z)
• Training on the Benchmark Is Not All You Need [52.01920740114261]
  We propose a simple and effective data leakage detection method based on the contents of multiple-choice options.<n>Our method is able to work under gray-box conditions without access to model training data or weights.<n>We evaluate the degree of data leakage of 35 mainstream open-source LLMs on four benchmark datasets.
  arXiv  Detail & Related papers  (2024-09-03T11:09:44Z)
• How to Train Data-Efficient LLMs [56.41105687693619]
  We study data-efficient approaches for pre-training language models (LLMs) We find that Ask-LLM and Density sampling are the best methods in their respective categories. In our comparison of 19 samplers, involving hundreds of evaluation tasks and pre-training runs, we find that Ask-LLM and Density are the best methods in their respective categories.
  arXiv  Detail & Related papers  (2024-02-15T02:27:57Z)
• DataComp: In search of the next generation of multimodal datasets [179.79323076587255]
  DataComp is a testbed for dataset experiments centered around a new candidate pool of 12.8 billion image-text pairs from Common Crawl. Our benchmark consists of multiple compute scales spanning four orders of magnitude. In particular, our best baseline, DataComp-1B, enables training a CLIP ViT-L/14 from scratch to 79.2% zero-shot accuracy on ImageNet.
  arXiv  Detail & Related papers  (2023-04-27T11:37:18Z)
• Fairness-Aware Data Valuation for Supervised Learning [4.874780144224057]
  We propose Fairness-Aware Data vauatiOn (FADO) to incorporate fairness concerns into a series of ML-related tasks. We show how FADO can be applied as the basis for unfairness mitigation pre-processing techniques. Our methods achieve promising results -- up to a 40 p.p. improvement in fairness at a less than 1 p.p. loss in performance compared to a baseline.
  arXiv  Detail & Related papers  (2023-03-29T18:51:13Z)

This list is automatically generated from the titles and abstracts of the papers in this site.

This site does not guarantee the quality of this site (including all information) and is not responsible for any consequences.