Corrective Diffusion Language Models
cs.LG
Submitted: 2025-12-17
Updated: 2026-11-05
Comments: NeurIPS 2026. 72 pages. Code: https://github.com/zhangshuibai/CDLM
Code: https://github.com/zhangshuibai/CDLM
License: http://arxiv.org/licenses/nonexclusive-distrib/1.0/
Terminology
Sources
- GPT-4 Technical Report
- Program Synthesis with Large Language Models
- Beyond Masked and Unmasked: Discrete Diffusion Models via Partial Masking
- Evaluating Large Language Models Trained on Code
- Optimal Inference Schedules for Masked Diffusion Models
- EDIT-Bench: Evaluating LLM Abilities to Perform Real-World Instructed Code Edits
- Gemini 2.5: Pushing the Frontier with Advanced Reasoning, Multimodality, Long Context, and Next Generation Agentic Capabilities
- The Llama 3 Herd of Models
- DeepSeek-R1: Incentivizing Reasoning Capability in LLMs via Reinforcement Learning
- Fast and Accurate Causal Parallel Decoding using Jacobi Forcing
- Accelerating Diffusion LLMs via Adaptive Parallel Decoding
- SWE-bench: Can Language Models Resolve Real-World GitHub Issues?
- ParallelBench: Understanding the Trade-offs of Parallel Decoding in Diffusion LLMs
- Fine-Tuning Masked Diffusion for Provable Self-Correction
- Mercury: Ultra-Fast Language Models Based on Diffusion
- Effective Test-Time Scaling of Discrete Diffusion through Iterative Refinement
- RefineBench: Evaluating Refinement Capability of Language Models via Checklists
- WeDLM: Reconciling Diffusion Language Models with Standard Causal Attention for Fast Inference
- TiDAR: Think in Diffusion, Talk in Autoregression
- NVIDIA Nemotron Nano 2: An Accurate and Efficient Hybrid Mamba-Transformer Reasoning Model
Related papers
- FlexMoE: Learning One-for-All Multi-Budget Expert Pruning for Flexible MoE Deployment
- OptiMind: Improving Small Language Models for Optimization Modeling through Formulation Guidance and Multi-Turn Learning
- Staying in the Sweet Spot: Boost Reasoning Evolution via Responsive Rollout Difficulty Manipulation
- Examining the Impact of Data and Training Methods on Emergent and Subliminal Misalignment
- Rethinking Black-Box On-Policy Distillation: A Rubric-based Analysis
- (Mis)generalization of Helpful-Only Fine-tuning