Deciphering the Impact of Pretraining Data on Large Language Models through Machine Unlearning

• URL: http://arxiv.org/abs/2402.11537v3
• Date: Wed, 28 Aug 2024 10:39:11 GMT
• Title: Deciphering the Impact of Pretraining Data on Large Language Models through Machine Unlearning
• Authors: Yang Zhao, Li Du, Xiao Ding, Kai Xiong, Zhouhao Sun, Jun Shi, Ting Liu, Bing Qin,
• Abstract summary: We systematically analyze the impact of 48 datasets from 5 major categories of pretraining data of Large Language Models. Our analyses provide empirical results about the contribution of multiple corpora on the performances of LLMs.
• Score: 45.96954837114004
• License: http://creativecommons.org/licenses/by/4.0/
• Abstract: Through pretraining on a corpus with various sources, Large Language Models (LLMs) have gained impressive performance. However, the impact of each component of the pretraining corpus remains opaque. As a result, the organization of the pretraining corpus is still empirical and may deviate from the optimal. To address this issue, we systematically analyze the impact of 48 datasets from 5 major categories of pretraining data of LLMs and measure their impacts on LLMs using benchmarks about nine major categories of model capabilities. Our analyses provide empirical results about the contribution of multiple corpora on the performances of LLMs, along with their joint impact patterns, including complementary, orthogonal, and correlational relationships. We also identify a set of ``high-impact data'' such as Books that is significantly related to a set of model capabilities. These findings provide insights into the organization of data to support more efficient pretraining of LLMs.

Related papers

• Data-Efficient Pretraining with Group-Level Data Influence Modeling [49.18903821780051]
  Group-Level Data Influence Modeling (Group-MATES) is a novel data-efficient pretraining method. Group-MATES collects oracle group-level influences by locally probing the pretraining model with data sets. It then fine-tunes a relational data influence model to approximate oracles as relationship-weighted aggregations of individual influences.
  arXiv  Detail & Related papers  (2025-02-20T16:34:46Z)
• Towards Better Understanding Table Instruction Tuning: Decoupling the Effects from Data versus Models [62.47618742274461]
  We fine-tune base models from the Mistral, OLMo, and Phi families on existing public training datasets. Our replication achieves performance on par with or surpassing existing table LLMs. We decouple the contributions of training data and the base model, providing insight into their individual impacts.
  arXiv  Detail & Related papers  (2025-01-24T18:50:26Z)
• Fine-tuning Large Language Models for Entity Matching [3.7277730514654555]
  Generative large language models (LLMs) are a promising alternative to pre-trained language models for entity matching. This paper explores the potential of fine-tuning LLMs for entity matching.
  arXiv  Detail & Related papers  (2024-09-12T16:20:57Z)
• Evaluating Linguistic Capabilities of Multimodal LLMs in the Lens of Few-Shot Learning [15.919493497867567]
  This study aims to evaluate the performance of Multimodal Large Language Models (MLLMs) on the VALSE benchmark. We conducted a comprehensive assessment of state-of-the-art MLLMs, varying in model size and pretraining datasets.
  arXiv  Detail & Related papers  (2024-07-17T11:26:47Z)
• Self-training Large Language Models through Knowledge Detection [26.831873737733737]
  Large language models (LLMs) often necessitate extensive labeled datasets and training compute to achieve impressive performance across downstream tasks. This paper explores a self-training paradigm, where the LLM autonomously curates its own labels and selectively trains on unknown data samples. Empirical evaluations demonstrate significant improvements in reducing hallucination in generation across multiple subjects.
  arXiv  Detail & Related papers  (2024-06-17T07:25:09Z)
• Unveiling the Impact of Coding Data Instruction Fine-Tuning on Large Language Models Reasoning [64.5243480989869]
  coding data is known to boost reasoning abilities during pretraining. Its role in activating internal reasoning capacities during IFT remains understudied. This paper investigates how coding data impact LLMs' reasoning capacities during IFT stage.
  arXiv  Detail & Related papers  (2024-05-30T23:20:25Z)
• To Repeat or Not To Repeat: Insights from Scaling LLM under Token-Crisis [50.31589712761807]
  Large language models (LLMs) are notoriously token-hungry during pre-training, and high-quality text data on the web is approaching its scaling limit for LLMs. We investigate the consequences of repeating pre-training data, revealing that the model is susceptible to overfitting. Second, we examine the key factors contributing to multi-epoch degradation, finding that significant factors include dataset size, model parameters, and training objectives.
  arXiv  Detail & Related papers  (2023-05-22T17:02:15Z)
• Measuring Causal Effects of Data Statistics on Language Model's `Factual' Predictions [59.284907093349425]
  Large amounts of training data are one of the major reasons for the high performance of state-of-the-art NLP models. We provide a language for describing how training data influences predictions, through a causal framework. Our framework bypasses the need to retrain expensive models and allows us to estimate causal effects based on observational data alone.
  arXiv  Detail & Related papers  (2022-07-28T17:36:24Z)

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.