English

PruneCD: Contrasting Pruned Self Model to Improve Decoding Factuality

Computation and Language 2025-09-24 v2 Artificial Intelligence

Abstract

To mitigate the hallucination problem in large language models, DoLa exploits early exit logits from the same model as a contrastive prior. However, we found that these early exit logits tend to be flat, low in magnitude, and fail to reflect meaningful contrasts. To address this, we propose PruneCD, a novel contrastive decoding method that constructs the amateur model via layer pruning rather than early exit. This design leads to more informative and well-aligned logits, enabling more effective contrastive decoding. Through qualitative and quantitative analyses, we demonstrate that PruneCD consistently improves factuality with minimal inference overhead, offering a robust and practical approach to mitigating hallucinations in LLMs.

Cite

@article{arxiv.2509.16598,
  title  = {PruneCD: Contrasting Pruned Self Model to Improve Decoding Factuality},
  author = {Byeongho Yu and Changhun Lee and Jungyu Jin and Eunhyeok Park},
  journal= {arXiv preprint arXiv:2509.16598},
  year   = {2025}
}

Comments

accepted at EMNLP 2025 Main Conference

R2 v1 2026-07-01T05:47:03.844Z