Holdout-Loss-Based Data Selection for LLM Finetuning via In-Context Learning

• URL: http://arxiv.org/abs/2510.14459v1
• Date: Thu, 16 Oct 2025 09:00:39 GMT
• Title: Holdout-Loss-Based Data Selection for LLM Finetuning via In-Context Learning
• Authors: Ling Zhang, Xianliang Yang, Juwon Yu, Park Cheonyoung, Lei Song, Jiang Bian,
• Abstract summary: We present a theoretically grounded, resource-efficient framework for data selection and reweighting.<n>At its core is an In-Context Approximation (ICA) that estimates the holdout loss a model would incur after training on a candidate example.<n>We derive per-example weights from ICA scores, dynamically reweighting gradient updates as model parameters evolve.
• Score: 19.677969862434708
• License: http://creativecommons.org/licenses/by/4.0/
• Abstract: Fine-tuning large pretrained language models is a common approach for aligning them with human preferences, but noisy or off-target examples can dilute supervision. While small, well-chosen datasets often match the performance of much larger ones, systematic and efficient ways to identify high-value training data remain underexplored. Many current methods rely on heuristics or expensive retraining. We present a theoretically grounded, resource-efficient framework for data selection and reweighting. At its core is an In-Context Approximation (ICA) that estimates the holdout loss a model would incur after training on a candidate example by conditioning on a small, curated holdout set in context. ICA requires no reference model and no additional finetuning. Under a local linearization, ICA is equivalent to a first-order update toward the holdout optimum, motivating its use as a proxy for data value. We derive per-example weights from ICA scores, dynamically reweighting gradient updates as model parameters evolve. Across SFT, DPO, and SimPO, and over diverse backbones and datasets, ICA-based reweighting consistently improves model alignment with minimal overhead. We analyze sensitivity to score update frequency and the choice of $k$ holdout examples for in-context demonstrations, and note limitations for rapidly drifting on-policy updates, highlighting directions for future work. Code and prompts will be released.

Related papers

• Entropy-Guided Token Dropout: Training Autoregressive Language Models with Limited Domain Data [89.96277093034547]
  We introduce EntroDrop, an entropy-guided token dropout method that functions as structured data regularization.<n>We show that EntroDrop consistently outperforms standard regularization baselines and maintains robust performance throughout extended multi-epoch training.
  arXiv  Detail & Related papers  (2025-12-29T12:35:51Z)
• Scalable Data Attribution via Forward-Only Test-Time Inference [3.5466521714943138]
  Data attribution seeks to trace model behavior back to the training examples that shaped it.<n>We propose a data attribution method that preserves the same first-order counterfactual target.<n>Our method provides a theoretical framework for practical, real-time data attribution in large pretrained models.
  arXiv  Detail & Related papers  (2025-11-25T00:11:39Z)
• Aligning Frozen LLMs by Reinforcement Learning: An Iterative Reweight-then-Optimize Approach [65.6966065843227]
  Iterative Reweight-then-IRO is a framework that performs RL-style alignment of a frozen base model without touching its parameters.<n>At test time, the value functions are used to guide the base model generation via a search-based optimization process.<n> Notably, users can apply IRO to align a model on their own dataset, similar to OpenAI's reinforcement fine-tuning (RFT)
  arXiv  Detail & Related papers  (2025-06-21T21:49:02Z)
• Maximally-Informative Retrieval for State Space Model Generation [59.954191072042526]
  We introduce Retrieval In-Context Optimization (RICO) to minimize model uncertainty for a particular query at test-time.<n>Unlike traditional retrieval-augmented generation (RAG), which relies on externals for document retrieval, our approach leverages direct feedback from the model.<n>We show that standard top-$k$ retrieval with model gradients can approximate our optimization procedure, and provide connections to the leave-one-out loss.
  arXiv  Detail & Related papers  (2025-06-13T18:08:54Z)
• ToReMi: Topic-Aware Data Reweighting for Dynamic Pre-Training Data Selection [28.75333303894706]
  ToReMi is a novel framework that adjusts training sample weights according to their topical associations and observed learning patterns.<n>Our experiments reveal that ToReMi variants consistently achieve superior performance over conventional pre-training approaches.
  arXiv  Detail & Related papers  (2025-04-01T12:06:42Z)
• Dynamic Loss-Based Sample Reweighting for Improved Large Language Model Pretraining [55.262510814326035]
  Existing reweighting strategies primarily focus on group-level data importance.<n>We introduce novel algorithms for dynamic, instance-level data reweighting.<n>Our framework allows us to devise reweighting strategies deprioritizing redundant or uninformative data.
  arXiv  Detail & Related papers  (2025-02-10T17:57:15Z)
• Context-aware Prompt Tuning: Advancing In-Context Learning with Adversarial Methods [69.36397993451742]
  This work introduces Context-aware Prompt Tuning (CPT), a method inspired by ICL, PT, and adversarial attacks. We modify specific context tokens, considering the unique structure of input and output formats. Inspired by adversarial attacks, we adjust the input based on the labels present in the context, focusing on minimizing, rather than maximizing, the loss.
  arXiv  Detail & Related papers  (2024-10-22T17:45:47Z)
• Few-Shot Load Forecasting Under Data Scarcity in Smart Grids: A Meta-Learning Approach [0.18641315013048293]
  This paper proposes adapting an established model-agnostic meta-learning algorithm for short-term load forecasting. The proposed method can rapidly adapt and generalize within any unknown load time series of arbitrary length. The proposed model is evaluated using a dataset of historical load consumption data from real-world consumers.
  arXiv  Detail & Related papers  (2024-06-09T18:59:08Z)
• Self-Augmented Preference Optimization: Off-Policy Paradigms for Language Model Alignment [104.18002641195442]
  We introduce Self-Augmented Preference Optimization (SAPO), an effective and scalable training paradigm that does not require existing paired data. Building on the self-play concept, which autonomously generates negative responses, we further incorporate an off-policy learning pipeline to enhance data exploration and exploitation.
  arXiv  Detail & Related papers  (2024-05-31T14:21:04Z)
• Fine-Tuning Pre-Trained Language Models Effectively by Optimizing Subnetworks Adaptively [32.001304911395756]
  We propose a Dynamic Selection (DPS) algorithm for the large-scale pre-trained models during fine-tuning. Experiments on the GLUE benchmark show that DPS outperforms previous fine-tuning methods in terms of overall performance and stability.
  arXiv  Detail & Related papers  (2022-11-03T08:32:12Z)

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.