diffusion language models
A class of probabilistic generative models used for text generation and understanding, leveraging diffusion processes to progressively refine and introduce noise into representations for better sampling quality.
- Anchored Diffusion Language Model
- Don’t Let It Fade: Preserving Edits in Diffusion Language Models via Token Timestep Allocation
- MRO: Enhancing Reasoning in Diffusion Language Models via Multi-Reward Optimization
- Reinforcing the Diffusion Chain of Lateral Thought with Diffusion Language Models
- STEAD: Robust Provably Secure Linguistic Steganography with Diffusion Language Model
- Theoretical Benefit and Limitation of Diffusion Language Model
- dKV-Cache: The Cache for Diffusion Language Models