Combatting Dimensional Collapse in LLM Pre-Training Data via Diversified File Selection

• URL: http://arxiv.org/abs/2504.20644v1
• Date: Tue, 29 Apr 2025 11:13:18 GMT
• Title: Combatting Dimensional Collapse in LLM Pre-Training Data via Diversified File Selection
• Authors: Ziqing Fan, Siyuan Du, Shengchao Hu, Pingjie Wang, Li Shen, Ya Zhang, Dacheng Tao, Yanfeng Wang,
• Abstract summary: DiverSified File selection algorithm (DiSF) is proposed to select the most decorrelated text files in the feature space.<n>DiSF saves 98.5% of 590M training files in SlimPajama, outperforming the full-data pre-training within a 50B training budget.
• Score: 65.96556073745197
• License: http://arxiv.org/licenses/nonexclusive-distrib/1.0/
• Abstract: Selecting high-quality pre-training data for large language models (LLMs) is crucial for enhancing their overall performance under limited computation budget, improving both training and sample efficiency. Recent advancements in file selection primarily rely on using an existing or trained proxy model to assess the similarity of samples to a target domain, such as high quality sources BookCorpus and Wikipedia. However, upon revisiting these methods, the domain-similarity selection criteria demonstrates a diversity dilemma, i.e.dimensional collapse in the feature space, improving performance on the domain-related tasks but causing severe degradation on generic performance. To prevent collapse and enhance diversity, we propose a DiverSified File selection algorithm (DiSF), which selects the most decorrelated text files in the feature space. We approach this with a classical greedy algorithm to achieve more uniform eigenvalues in the feature covariance matrix of the selected texts, analyzing its approximation to the optimal solution under a formulation of $\gamma$-weakly submodular optimization problem. Empirically, we establish a benchmark and conduct extensive experiments on the TinyLlama architecture with models from 120M to 1.1B parameters. Evaluating across nine tasks from the Harness framework, DiSF demonstrates a significant improvement on overall performance. Specifically, DiSF saves 98.5% of 590M training files in SlimPajama, outperforming the full-data pre-training within a 50B training budget, and achieving about 1.5x training efficiency and 5x data efficiency.

Related papers

• SHA256 at SemEval-2025 Task 4: Selective Amnesia -- Constrained Unlearning for Large Language Models via Knowledge Isolation [12.838593066237452]
  Large language models (LLMs) memorize frequently sensitive information during training, posing risks when deploying publicly accessible models.<n>This paper presents our solution to SemEval-2025 Task 4 on targeted unlearning, which combines causal mediation analysis with layer-specific optimization.
  arXiv  Detail & Related papers  (2025-04-17T15:05:40Z)
• Data Mixture Optimization: A Multi-fidelity Multi-scale Bayesian Framework [10.317740844867913]
  We build a simulator based on 472 language model pre-training runs with varying data compositions from the SlimPajama dataset.<n>We observe that even simple acquisition functions can enable principled training decisions across training models from 20M to 1B kernels.
  arXiv  Detail & Related papers  (2025-03-26T22:19:47Z)
• MM-RLHF: The Next Step Forward in Multimodal LLM Alignment [59.536850459059856]
  We introduce MM-RLHF, a dataset containing $mathbf120k$ fine-grained, human-annotated preference comparison pairs.<n>We propose several key innovations to improve the quality of reward models and the efficiency of alignment algorithms.<n>Our approach is rigorously evaluated across $mathbf10$ distinct dimensions and $mathbf27$ benchmarks.
  arXiv  Detail & Related papers  (2025-02-14T18:59:51Z)
• Dividable Configuration Performance Learning [4.949726352498762]
  We propose a model-agnostic and sparsity-robust framework for predicting configuration performance, dubbed DaL. DaL is based on the new paradigm of dividable learning that builds a model via "divide-and-learn"
  arXiv  Detail & Related papers  (2024-09-11T21:23:23Z)
• Landscape-Aware Automated Algorithm Configuration using Multi-output Mixed Regression and Classification [0.01649298969786889]
  We investigate the potential of randomly generated functions (RGF) for the model training. We focus on automated algorithm configuration (AAC) We analyze the performance of dense neural network (NN) models in handling the mixed regression and classification tasks.
  arXiv  Detail & Related papers  (2024-09-02T20:04:41Z)
• Clustering and Ranking: Diversity-preserved Instruction Selection through Expert-aligned Quality Estimation [56.13803674092712]
  We propose an industrial-friendly, expert-aligned and diversity-preserved instruction data selection method: Clustering and Ranking (CaR) CaR employs a two-step process: first, it ranks instruction pairs using a high-accuracy (84.25%) scoring model aligned with expert preferences; second, it preserves dataset diversity through clustering. In our experiment, CaR efficiently selected a mere 1.96% of Alpaca's IT data, yet the resulting AlpaCaR model surpassed Alpaca's performance by an average of 32.1% in GPT-4 evaluations.
  arXiv  Detail & Related papers  (2024-02-28T09:27:29Z)
• 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)
• Efficient Grammatical Error Correction Via Multi-Task Training and Optimized Training Schedule [55.08778142798106]
  We propose auxiliary tasks that exploit the alignment between the original and corrected sentences. We formulate each task as a sequence-to-sequence problem and perform multi-task training. We find that the order of datasets used for training and even individual instances within a dataset may have important effects on the final performance.
  arXiv  Detail & Related papers  (2023-11-20T14:50:12Z)
• Self-Directed Online Machine Learning for Topology Optimization [58.920693413667216]
  Self-directed Online Learning Optimization integrates Deep Neural Network (DNN) with Finite Element Method (FEM) calculations. Our algorithm was tested by four types of problems including compliance minimization, fluid-structure optimization, heat transfer enhancement and truss optimization. It reduced the computational time by 2 5 orders of magnitude compared with directly using methods, and outperformed all state-of-the-art algorithms tested in our experiments.
  arXiv  Detail & Related papers  (2020-02-04T20:00:28Z)

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.