paper-with-me

Papers

FOCUS & RePAIR: Mitigating Text Degeneration via Token-Level Guidance for Pruned Large Language Models

2026-08-27 · Junyoung Lee, Sehyeon Park, Shinhyoung Jang, Seonha Ryu, Hojeong Kim, Hyunsei Lee, Il Hong Suh, Yeseong Kim arxiv

Pruning is a practical approach to compress large language models (LLMs), but it can amplify text degeneration, especially repetition loops, even when perplexity and task accuracy remain largely unchanged. In this work, we present a token-level analysis of this failure mode by viewing decoding as a dynamical process that enters and persists in a small set of recurrent contexts. Our analysis decomposes degeneration into loop entry risk and loop persistence, and shows that persistence is controlled by the escape mass assigned to plausible alternatives within the token sampling set. Motivated by these findings, we propose two token-level guidance objectives for post-pruning fine-tuning. FOCUS reweights distillation toward high-confidence teacher regions to suppress leakage, while RePAIR uses onset-centered positive/negative continuation pairs with a margin loss to promote plausible alternatives and prevent early commitment to repetition loops. Experiments on open-ended continuation and instruction-based generation show that both methods consistently reduce repetition and improve generation quality.

📄 PDF Abstract BibTeX arXiv:2608.26676

Code (2)

Tavish9/awesome-daily-AI-arxiv ★ 113
arxivsub/arXivSub_daily_arxiv ★ 4

Similar Papers 제목 키워드 기반

Rare Tokens Degenerate All Tokens: Improving Neural Text Generation via Adaptive Gradient Gating for Rare Token Embeddings

2021-11-16 · ACL ARR November 2021 11 · Anonymous

Recent studies have determined that the learned token embeddings of large-scale neural language models are degenerated to be anisotropic with a narrow-cone shape. This phenomenon, called the representation degeneration p…

AllLanguage ModelingLanguage ModellingMachine Translation+2

Rare Tokens Degenerate All Tokens: Improving Neural Text Generation via Adaptive Gradient Gating for Rare Token Embeddings

2021-09-07 · ACL 2022 5 · Sangwon Yu, Jongyoon Song, Heeseung Kim, Seong-min Lee 외

Recent studies have determined that the learned token embeddings of large-scale neural language models are degenerated to be anisotropic with a narrow-cone shape. This phenomenon, called the representation degeneration p…

AllLanguage ModelingLanguage ModellingMachine Translation+3

A Simple Contrastive Learning Objective for Alleviating Neural Text Degeneration

2022-05-05 · Shaojie Jiang, Ruqing Zhang, Svitlana Vakulenko, Maarten de Rijke

The cross-entropy objective has proved to be an all-purpose training objective for autoregressive language models (LMs). However, without considering the penalization of problematic tokens, LMs trained using cross-entrop…

Contrastive LearningDialogue GenerationLanguage ModelingLanguage Modelling

Mitigating Data Imbalance and Representation Degeneration in Multilingual Machine Translation

2023-05-22 · Wen Lai, Alexandra Chronopoulou, Alexander Fraser

Despite advances in multilingual neural machine translation (MNMT), we argue that there are still two major challenges in this area: data imbalance and representation degeneration. The data imbalance problem refers to th…

Contrastive LearningMachine TranslationTranslation

Sanitize Your Responses: Mitigating Privacy Leakage in Large Language Models

2025-09-29 · Wenjie Fu, Huandong Wang, Junyao Gao, Guoan Wan 외 arxiv

As Large Language Models (LLMs) achieve remarkable success across a wide range of applications, such as chatbots and code copilots, concerns surrounding the generation of harmful content have come increasingly into focus…