An Empirical Study on Influence-Based Pretraining Data Selection for Code Large Language Models

• URL: http://arxiv.org/abs/2604.07769v1
• Date: Thu, 09 Apr 2026 03:48:11 GMT
• Title: An Empirical Study on Influence-Based Pretraining Data Selection for Code Large Language Models
• Abstract summary: This paper explores the effectiveness of data-influence-score filtering within the context of programming-related datasets.<n>We pre-train a Code-LLMs with 1 billion parameters from scratch on a dataset of 100 billion code tokens.<n>Our findings show that data-influence-score filtering based on validation-set-loss can enhance models programming performance.
• Score: 30.70386124413396
• License: http://arxiv.org/licenses/nonexclusive-distrib/1.0/
• Abstract: Recent advancements in code large language models (Code-LLMs) have demonstrated remarkable capabilities in resolving programming related tasks. Meanwhile, researchers have recognized that the quality of pre-training data is crucial for improving LLM performance. However, most of the existing research on pre-training data filtering has focused on general datasets, and little attention for programming datasets. In this paper, we aim to address this gap by exploring the effectiveness of a widely used general data filtering technique, i.e., data-influence-score filtering, within the context of programming-related datasets. To this end, we first introduce a method for calculating data-influence-score for generative programming tasks which involves transforming a variety of downstream coding tasks into validation sets and using the models loss on these sets as a performance metric. Next, we pre-train a Code-LLMs with 1 billion parameters from scratch on a dataset of 100 billion code tokens. Based on it, we conduct an extensive empirical study to evaluate the effectiveness of data-influence-score filtering methods. Specifically, we examine how well this technique improves model performance, investigate how the characteristics of beneficial training data vary across different training stages and programming tasks, and assess the feasibility of prediction-based data-influence-score filtering method. Our findings show that data-influence-score filtering based on validation-set-loss can enhance models programming performance. Moreover, we observe that the criteria of beneficial training data differ significantly across various downstream programming tasks.

Related papers

• Influcoder: Distilling Decoders' Gradient Influence Rankings into an Encoder for Data Attribution [3.1857077981565793]
  Data Attribution methods estimate how samples in a training dataset can precondition a model to generate certain outputs.<n>We propose a method, Influcoder, as a quick and cost-effective approach to influence-based Data Attribution at scale.
  arXiv  Detail & Related papers  (2026-06-11T17:58:33Z)
• Learning Multi-Indicator Weights for Data Selection: A Joint Task-Model Adaptation Framework with Efficient Proxies [50.39041754816285]
  We propose a framework for learning multi-validation weights that jointly adapts data selection to both the downstream task and the specific model.<n>Our approach achieves performance comparable to or exceeding full-dataset tuning while using only 30% of the training samples on GSM8K.
  arXiv  Detail & Related papers  (2026-05-10T17:30:16Z)
• Data Efficacy for Language Model Training [29.901090317084005]
  Data is fundamental to the training of language models (LM)<n>Recent research has been dedicated to data efficiency, which aims to maximize performance by selecting a minimal or optimal subset of training data.<n>This work introduces a general paradigm, DELT, for considering data efficacy in LM training.
  arXiv  Detail & Related papers  (2025-06-26T17:59:07Z)
• DataMIL: Selecting Data for Robot Imitation Learning with Datamodels [77.48472034791213]
  We introduce DataMIL, a policy-driven data selection framework built on the datamodels paradigm.<n>Unlike standard practices that filter data using human notions of quality, DataMIL directly optimize data selection for task success.<n>We validate our approach on a suite of more than 60 simulation and real-world manipulation tasks.
  arXiv  Detail & Related papers  (2025-05-14T17:55:10Z)
• OpenCodeReasoning: Advancing Data Distillation for Competitive Coding [61.15402517835137]
  We build a supervised fine-tuning (SFT) dataset to achieve state-of-the-art coding capability results in models of various sizes.<n>Our models use only SFT to achieve 61.8% on LiveCodeBench and 24.6% on CodeContests, surpassing alternatives trained with reinforcement learning.
  arXiv  Detail & Related papers  (2025-04-02T17:50:31Z)
• Capturing the Temporal Dependence of Training Data Influence [100.91355498124527]
  We formalize the concept of trajectory-specific leave-one-out influence, which quantifies the impact of removing a data point during training.<n>We propose data value embedding, a novel technique enabling efficient approximation of trajectory-specific LOO.<n>As data value embedding captures training data ordering, it offers valuable insights into model training dynamics.
  arXiv  Detail & Related papers  (2024-12-12T18:28:55Z)
• ROSE: A Reward-Oriented Data Selection Framework for LLM Task-Specific Instruction Tuning [26.069986067128912]
  We introduce ROSE, a Reward-Oriented inStruction data sElection method to optimize data selection for task-specific instruction tuning.<n> ROSE adapts an influence formulation to approximate the influence of training data points relative to a few-shot preference validation set to select the most task-related training data points.
  arXiv  Detail & Related papers  (2024-12-01T01:01:09Z)
• LESS: Selecting Influential Data for Targeted Instruction Tuning [64.78894228923619]
  We propose LESS, an efficient algorithm to estimate data influences and perform Low-rank gradiEnt Similarity Search for instruction data selection. We show that training on a LESS-selected 5% of the data can often outperform training on the full dataset across diverse downstream tasks. Our method goes beyond surface form cues to identify data that the necessary reasoning skills for the intended downstream application.
  arXiv  Detail & Related papers  (2024-02-06T19:18:04Z)
• DsDm: Model-Aware Dataset Selection with Datamodels [81.01744199870043]
  Standard practice is to filter for examples that match human notions of data quality. We find that selecting according to similarity with "high quality" data sources may not increase (and can even hurt) performance compared to randomly selecting data. Our framework avoids handpicked notions of data quality, and instead models explicitly how the learning process uses train datapoints to predict on the target tasks.
  arXiv  Detail & Related papers  (2024-01-23T17:22:00Z)
• One-Shot Learning as Instruction Data Prospector for Large Language Models [108.81681547472138]
  textscNuggets uses one-shot learning to select high-quality instruction data from extensive datasets. We show that instruction tuning with the top 1% of examples curated by textscNuggets substantially outperforms conventional methods employing the entire dataset.
  arXiv  Detail & Related papers  (2023-12-16T03:33:12Z)
• Data-Centric Machine Learning in the Legal Domain [0.2624902795082451]
  This paper explores how changes in a data set influence the measured performance of a model. Using three publicly available data sets from the legal domain, we investigate how changes to their size, the train/test splits, and the human labelling accuracy impact the performance. The observed effects are surprisingly pronounced, especially when the per-class performance is considered.
  arXiv  Detail & Related papers  (2022-01-17T23:05:14Z)

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.