Review, Remask, Refine (R3): Process-Guided Block Diffusion for Text Generation
arxiv.org · 3,532 words · saved by 1 readers
N/A
Review, Remask, Refine (R3): Process-Guided Block Diffusion for Text Generation Nikita Mounier * 1 Parsa Ideahpour * 2 Abstract A key challenge for iterative text generation is enabling models to efficiently identify and correct their own errors. arXiv:2507.08018v1 [cs.CL] 7 Jul 2025…
related reading
- Large Language Diffusion Modelsarxiv.org
- Effective Test-Time Scaling of Discrete Diffusion through Iterative Refinementarxiv.org
- Illustrating Reinforcement Learning from Human Feedback (RLHF)huggingface.co
- 2503.09573arxiv.org
- 2409.02908arxiv.org
- Kuleshov Group | How to Build a Diffusion Language Modelkuleshov-group.github.io
- Modifying LLM Beliefs with Synthetic Document Finetuningalignment.anthropic.com
- 2406.07524arxiv.org
- Lookahead Unmasking Elicits Accurate Decoding in Diffusion Language Modelsarxiv.org
- Adaptive Classifier-Free Guidance via Dynamic Low-Confidence Maskingarxiv.org
- Diffusion will be everywherevarunneal.github.io
- Efficient Guided Generation for Large Language Modelsarxiv.org