diffusion-language-models

Tag

Cards List
#diffusion-language-models

Multi-Token Residual Prediction

arXiv cs.LG ↗ · 2026-05-20

Introduces Multi-token Residual Prediction (MRP), a lightweight module for diffusion language models that enables dependency-aware multi-token denoising within a single backbone forward pass, achieving up to 1.42× lossless speedup.

0 favorites 0 likes
#diffusion-language-models

@NVIDIAAI: Most language models only generate one token at a time. We just released Nemotron-Labs-Diffusion, a family of diffusion…

X AI KOLs Following ↗ · 2026-05-19 Cached

NVIDIA released Nemotron-Labs-Diffusion, a family of diffusion language models that generate multiple tokens in parallel, enabling faster inference and better GPU utilization, with sizes from 3B to 14B including vision-language variants.

0 favorites 0 likes
#diffusion-language-models

DACA-GRPO: Denoising-Aware Credit Assignment for Reinforcement Learning in Diffusion Language Models

arXiv cs.LG ↗ · 2026-05-19 Cached

This paper identifies weaknesses in existing reinforcement learning methods for diffusion language models—lack of temporal credit assignment and biased likelihood estimates—and proposes DACA-GRPO, a plug-and-play enhancement that introduces denoising progress scores and stratified masking likelihood, achieving consistent improvements across reasoning, code generation, and constrained generation benchmarks.

0 favorites 0 likes
#diffusion-language-models

Dynamic Chunking for Diffusion Language Models

arXiv cs.CL ↗ · 2026-05-18 Cached

This paper introduces Dynamic Chunking for Diffusion Language Models (DCDM), which replaces fixed positional blocks in block discrete diffusion with content-defined semantic chunks using a differentiable Chunking Attention mechanism, achieving consistent improvements across scales up to 1.5B parameters.

0 favorites 0 likes
#diffusion-language-models

@DailyDoseOfDS_: Turn any Autoregressive LLM into a Diffusion LM. dLLM is a Python library that unifies the training & evaluation of dif…

X AI KOLs Timeline ↗ · 2026-05-16 Cached

dLLM is an open-source Python library that allows converting any autoregressive language model into a diffusion language model with minimal compute, unifying training and evaluation.

0 favorites 0 likes
#diffusion-language-models

Beyond Mode-Seeking RL: Trajectory-Balance Post-Training for Diffusion Language Models

arXiv cs.LG ↗ · 2026-05-15 Cached

This paper identifies a failure mode called 'trajectory locking' in reward-maximizing post-training for diffusion language models, and proposes TraFL, a trajectory-balance objective that improves diversity and performance across math and code benchmarks.

0 favorites 0 likes
#diffusion-language-models

LEAP: Unlocking dLLM Parallelism via Lookahead Early-Convergence Token Detection

arXiv cs.LG ↗ · 2026-05-13 Cached

This paper introduces LEAP, a training-free method to accelerate inference in Diffusion Language Models (dLLMs) by detecting early-converging tokens, reducing denoising steps by 30% without losing accuracy.

0 favorites 0 likes
#diffusion-language-models

Self-Distilled Trajectory-Aware Boltzmann Modeling: Bridging the Training-Inference Discrepancy in Diffusion Language Models

arXiv cs.CL ↗ · 2026-05-13 Cached

This paper introduces TABOM, a self-distilled trajectory-based post-training framework for Diffusion Language Models that aligns training with inference trajectories using Boltzmann modeling to mitigate the training-inference discrepancy and reduce catastrophic forgetting.

0 favorites 0 likes
#diffusion-language-models

Don't Retrain, Align: Adapting Autoregressive LMs to Diffusion LMs via Representation Alignment

arXiv cs.LG ↗ · 2026-05-11 Cached

This paper introduces Repr-Align, a method to adapt autoregressive language models into diffusion language models via representation alignment, achieving up to 4x training acceleration without retraining representations from scratch.

0 favorites 0 likes
#diffusion-language-models

DiffRetriever: Parallel Representative Tokens for Retrieval with Diffusion Language Models

Hugging Face Daily Papers ↗ · 2026-05-08 Cached

This paper introduces DiffRetriever, a method that uses diffusion language models to generate multiple representative tokens in parallel for efficient information retrieval, outperforming autoregressive baselines in speed and accuracy.

0 favorites 0 likes
#diffusion-language-models

DALM: A Domain-Algebraic Language Model via Three-Phase Structured Generation

arXiv cs.CL ↗ · 2026-04-20 Cached

DALM proposes a domain-algebraic language model that generates text under exact structural constraints derived from a domain lattice, addressing hallucination by organizing knowledge into separate domain fibers with algebraic guarantees. The model uses three-phase structured denoising (domain → relation → concept) with domain-annotated training data to prevent cross-domain contamination.

0 favorites 0 likes
← Previous
← Back to home

Submit Feedback