Rethinking Soft Tokens for Parallel Decoding in Diffusion Language Models
cs.LG
Submitted: 2026-09-29
Updated: 2026-09-29
Code: https://github.com/kodaikawamura/rethinking-soft-tokens
Sources
- Program Synthesis with Large Language Models
- LLaDA2.0: Scaling Up Diffusion Language Models to 100B
- Evaluating Large Language Models Trained on Code
- DMax: Aggressive Parallel Decoding for dLLMs
- Training Verifiers to Solve Math Word Problems
- Breaking the Factorization Barrier in Diffusion Language Models
- Lost in Interpolation: Why Predictive Feedback Fails in Diffusion Language Models
- SoftCoT++: Test-Time Scaling with Soft Chain-of-Thought Reasoning
- Dream 7B: Diffusion Large Language Models
- dLLM: Simple Diffusion Language Modeling
Related papers
- Polynomial-Augmented Neural Networks (PANNs) with Weak Orthogonality Constraints for Enhanced Function and PDE Approximation
- AIRL-S: Unifying Reinforcement Learning and Search-Based Test-Time Scaling via Adversarial Inverse Reinforcement Learning
- Transformers as Bayesian In-Context Experimenters: Smoothness-Adaptive Efficient ATE Estimation
- Convergence issues in Relational Concept Analysis based on AOC-posets
- Beliefs Beyond Posteriors: The Predictive Cost of ELBO Binding
- Understanding Diffusion Models via Ratio-Based Function Approximation with SignReLU Networks