paper-with-me

홈 › Papers

Low-Confidence Gold: Refining Low-Confidence Samples for Efficient Instruction Tuning

2025-02-26 · Hongyi Cal, Jie Li, Wenzhen Dong

The effectiveness of instruction fine-tuning for Large Language Models is fundamentally constrained by the quality and efficiency of training datasets. This work introduces Low-Confidence Gold (LCG), a novel filtering framework that employs centroid-based clustering and confidence-guided selection for identifying valuable instruction pairs. Through a semi-supervised approach using a lightweight classifier trained on representative samples, LCG curates high-quality subsets while preserving data diversity. Experimental evaluation demonstrates that models fine-tuned on LCG-filtered subsets of 6K samples achieve superior performance compared to existing methods, with substantial improvements on MT-bench and consistent gains across comprehensive evaluation metrics. The framework's efficacy while maintaining model performance establishes a promising direction for efficient instruction tuning.

📄 PDF Abstract BibTeX arXiv:2502.18978

Code (0)

등록된 구현이 없습니다.

Tasks

Diversity

Similar Papers 제목 키워드 기반

Learning Robust Pair Confidence for Multimodal Emotion-Cause Pair Extraction

2026-06-17 · Zhuangzhuang Pan, Ning Dong, Yingna Su, Yan Xia arxiv

Multimodal emotion-cause pair extraction (MECPE) requires reliable pair confidence over candidate pairs. Existing pair scorers commonly use pair-level cross entropy over valid candidates, which treats links mostly indepe…

Emotion-Cause Pair Extraction

Typicalness-Aware Learning for Failure Detection

2024-11-04 · Yijun Liu, Jiequan Cui, Zhuotao Tian, Senqiao Yang 외

Deep neural networks (DNNs) often suffer from the overconfidence issue, where incorrect predictions are made with high confidence scores, hindering the applications in critical systems. In this paper, we propose a novel …

Deconstructing the Goldilocks Zone of Neural Network Initialization

2024-02-05 · Artem Vysogorets, Anna Dawid, Julia Kempe

The second-order properties of the training loss have a massive impact on the optimization dynamics of deep learning models. Fort & Scherlis (2019) discovered that a large excess of positive curvature and local convexity…

Unleashing the Potential of All Test Samples: Mean-Shift Guided Test-Time Adaptation

2025-07-01 · Jizhou Han, Chenhao Ding, SongLin Dong, Yuhang He 외 arxiv

Visual-language models (VLMs) like CLIP exhibit strong generalization but struggle with distribution shifts at test time. Existing training-free test-time adaptation (TTA) methods operate strictly within CLIP's original …

Test-time Adaptation

Are You Sure You're Sure? On the Impact of Instruction Tuning on Confidence and Lexical Diversity

2026-08-13 · Irina Proskurina, Mayank Kumar, Oyindolapo O. Komolafe arxiv

Instruction-tuned language models achieve strong performance across a range of generation tasks, but have also recently been shown to exhibit verbalized overconfidence. In question answering, verbalized model overconfide…

Question AnsweringAnswer Selection