# DataMan: Data Manager for Pre-training Large Language Models

arXiv: https://arxiv.org/abs/2502.19363 · HTML: https://mlsanity.com/paper/2502.19363/

## Key facts

- Cited by 17 papers in the map (data through 14 Oct 2026).
- First posted on arXiv 26 Feb 2025.

## Built on by (most cited first, 17 of 17)

- [RefineX: Learning to Refine Pre-training Data at Scale from Expert-Guided Programs](https://mlsanity.com/paper/2507.03253/) (2025)
- [RePro: Training Language Models to Faithfully Recycle the Web for Pretraining](https://mlsanity.com/paper/2510.10681/) (2025)
- [propella-1: Multi-Property Document Annotation for LLM Data Curation at Scale](https://mlsanity.com/paper/2602.12414/) (2026)
- [AttentionInfluence: Adopting Attention Head Influence for Weak-to-Strong Pretraining Data Selection](https://mlsanity.com/paper/2505.07293/) (2025)
- [MuRating: A High Quality Data Selecting Approach to Multilingual Large Language Model Pretraining](https://mlsanity.com/paper/2507.01785/) (2025)
- [ADEPT: Continual Pretraining via Adaptive Expansion and Dynamic Decoupled Tuning](https://mlsanity.com/paper/2510.10071/) (2025)
- [Bagpiper: Solving Open-Ended Audio Tasks via Rich Captions](https://mlsanity.com/paper/2602.05220/) (2026)
- [AIR: Post-training Data Selection for Reasoning via Attention Head Influence](https://mlsanity.com/paper/2512.13279/) (2025)
- [Generating Pretraining Tokens from Organic Data for Data-Bound Scaling](https://mlsanity.com/paper/2605.17849/) (2026)
- [DeMix: Debugging Training Data with Mixed Data Error Types by Investigating Influence Vectors](https://mlsanity.com/paper/2606.11616/) (2026)
- [DataOrchestra: Learning to Orchestrate Per-Example Curation of Pretraining Data](https://mlsanity.com/paper/2607.24717/) (2026)
- [Estimating the Effects of Sample Training Orders for Large Language Models without Retraining](https://mlsanity.com/paper/2505.22042/) (2025)
- [Less Data, Faster Convergence: Goal-Driven Data Optimization for Multimodal Instruction Tuning](https://mlsanity.com/paper/2603.12478/) (2026)
- [Harnessing the Potential of Optimizing Data Mixtures via Bayesian Domain Reweighting](https://mlsanity.com/paper/2607.27928/) (2026)
- [Towards Scalable Data Diversification for Language Model Pretraining via Leverage Score Sampling](https://mlsanity.com/paper/2609.32484/) (2026)
- [ReScraper: Unified Scraping and Cleaning of Web Data for Effective LLM Pretraining](https://mlsanity.com/paper/2609.34287/) (2026)
- [Fisher-Guided Submodular Data Selection for Continual Pre-Training of Large Language Models](https://mlsanity.com/paper/2610.02593/) (2026)

## Abstract (arXiv)

The performance emergence of large language models (LLMs) driven by data scaling laws makes the selection of pre-training data increasingly important. However, existing methods rely on limited heuristics and human intuition, lacking comprehensive and clear guidelines. To address this, we are inspired by ``reverse thinking'' -- prompting LLMs to self-identify which criteria benefit its performance. As its pre-training capabilities are related to perplexity (PPL), we derive 14 quality criteria from the causes of text perplexity anomalies and introduce 15 common application domains to support domain mixing. In this paper, we train a Data Manager (DataMan) to learn quality ratings and domain recognition from pointwise rating, and use it to annotate a 447B token pre-training corpus with 14 quality ratings and domain type. Our experiments validate our approach, using DataMan to select 30B tokens to train a 1.3B-parameter language model, demonstrating significant improvements in in-context learning (ICL), perplexity, and instruction-following ability over the state-of-the-art baseline. The best-performing model, based on the Overall Score l=5 surpasses a model trained with 50% more data using uniform sampling. We continue pre-training with high-rated, domain-specific data annotated by DataMan to enhance domain-specific ICL performance and thus verify DataMan's domain mixing ability. Our findings emphasize the importance of quality ranking, the complementary nature of quality criteria, and their low correlation with perplexity, analyzing misalignment between PPL and ICL performance. We also thoroughly analyzed our pre-training dataset, examining its composition, the distribution of quality ratings, and the original document sources.