Paper Detail
Wenlong Deng, Qi Zeng, Jiaming Zhang, Minghui Chen, Zixin Ding, Christos Thrampoulidis, Boying Gong, Xiaoxiao Li
Data valuation is essential for enhancing the transparency and accountability of large language models (LLMs) and vision-language models (VLMs). However, existing methods typically rely on gradient computations, making them computationally prohibitive for billion-parameter models and precluding batch parallelization. In this work, we introduce For-Value, a forward-only data valuation framework that enables efficient batch-scalable value estimation while maintaining effectiveness. Leveraging the expressive power of pretrained LLMs/VLMs, we theoretically demonstrate that data valuation can be captured by the alignment between the final hidden representations and prediction errors at the last layer. In light of this insight, For-Value computes data value using a simple closed-form expression with a single forward pass, eliminating the need for costly backpropagation and enabling efficient batch calculating at scale. Extensive experiments show that For-Value matches or outperforms gradient-based baselines in detecting influential data and mislabeled data, while achieving significant efficiency improvements.
No structured notes yet. Add `summary_sections`, `why_relevant`, `claim_impact`, or `next_action` in `papers.jsonl` to enrich this view.
No ranking explanation is available yet.
No tags.
@misc{deng2026value,
title = {For-Value: Efficient Forward-Only Data Valuation for finetuning LLMs and VLMs},
author = {Wenlong Deng and Qi Zeng and Jiaming Zhang and Minghui Chen and Zixin Ding and Christos Thrampoulidis and Boying Gong and Xiaoxiao Li},
year = {2026},
abstract = {Data valuation is essential for enhancing the transparency and accountability of large language models (LLMs) and vision-language models (VLMs). However, existing methods typically rely on gradient computations, making them computationally prohibitive for billion-parameter models and precluding batch parallelization. In this work, we introduce For-Value, a forward-only data valuation framework that enables efficient batch-scalable value estimation while maintaining effectiveness. Leveraging the },
url = {https://huggingface.co/papers/2508.10180},
keywords = {data valuation, large language models, vision-language models, gradient computations, forward-only, batch-scalable, hidden representations, prediction errors, closed-form expression, single forward pass, gradient-based baselines, code available, huggingface daily},
eprint = {2508.10180},
archiveprefix = {arXiv},
}
{}