paper-with-me

Papers

Self-Generated Error Training for Token Editing in Diffusion Language Models

2026-06-15 · Lin Yao arxiv

Token-to-token (T2T) editing lets LLaDA2.1 revise committed tokens during block-diffusion decoding. The released recipe trains this editor on random vocabulary corruptions, but at inference the editor sees the model's own fluent, high-confidence draft errors instead. We study this training-inference mismatch and propose self-generated T2T, which performs a no-gradient draft pass, fills masked positions with predicted tokens, and supervises recovery in a second pass under these self-generated corruptions. We implement the update as a short LoRA continued-pretraining pass on LLaDA2.1-mini and evaluate on several benchmarks under the official Q-Mode T2T procedure with unchanged inference parameters. The method generally improves accuracy while reducing T2T edit intensity, mitigating failure modes such as final-digit transcription errors after otherwise correct reasoning and excessive self-correction before short factual answers.

📄 PDF Abstract BibTeX arXiv:2606.17175

Code (0)

등록된 구현이 없습니다.

Similar Papers 제목 키워드 기반

How to Synthesize Text Data without Model Collapse?

2024-12-19 · Xuekai Zhu, Daixuan Cheng, Hengli Li, Kaiyan Zhang 외

Model collapse in synthetic data indicates that iterative training on self-generated data leads to a gradual decline in performance. With the proliferation of AI models, synthetic data will fundamentally reshape the web …

Targeted Remasking: Replacing Token Editing with Token-to-Mask Refinement in Discrete Diffusion Language Models

2026-04-06 · Lin Yao arxiv

Discrete masked diffusion language models such as LLaDA generate text through iterative denoising, where mask tokens are progressively replaced with predicted tokens. LLaDA2.1 introduced a Token-to-Token (T2T) editing me…

Instruction Following

VidToMe: Video Token Merging for Zero-Shot Video Editing

2023-12-17 · CVPR 2024 1 · Xirui Li, Chao Ma, Xiaokang Yang, Ming-Hsuan Yang

Diffusion models have made significant advances in generating high-quality images, but their application to video generation has remained challenging due to the complexity of temporal motion. Zero-shot video editing offe…

Video EditingVideo Generation

Subtle Errors Matter: Preference Learning via Error-injected Self-editing

2024-10-09 · Kaishuai Xu, Tiezheng Yu, Wenjun Hou, Yi Cheng 외

Large Language Models (LLMs) have exhibited strong mathematical reasoning and computational prowess, tackling tasks ranging from basic arithmetic to advanced competition-level problems. However, frequently occurring subt…

GSM8KMathMathematical Reasoning

Anchor Token Matching: Implicit Structure Locking for Training-free AR Image Editing

2025-04-14 · Taihang Hu, Linxuan Li, Kai Wang, Yaxing Wang 외

Text-to-image generation has seen groundbreaking advancements with diffusion models, enabling high-fidelity synthesis and precise image editing through cross-attention manipulation. Recently, autoregressive (AR) models h…

Image GenerationText to Image GenerationText-to-Image Generation