diffusion LLM — 3 Technical Articles

Machine Learning Algorithms & Natural Language Processing

Mar 3, 2026 · Artificial Intelligence

Enabling Search Agents to Think While Waiting: Diffusion LLMs Deliver 15% Faster Inference Without Accuracy Loss

The paper introduces DLLM‑Searcher, which equips diffusion large language models with a two‑stage training pipeline and a P‑ReAct inference scheme, allowing the model to issue tool calls while simultaneously reasoning, yielding 14‑22% end‑to‑end speedup and matching or surpassing traditional autoregressive agents on multi‑hop QA benchmarks.

Multi-hop QAP-ReActagentic training

0 likes · 10 min read

Enabling Search Agents to Think While Waiting: Diffusion LLMs Deliver 15% Faster Inference Without Accuracy Loss

Data Party THU

Oct 10, 2025 · Artificial Intelligence

How DPad Cuts Inference Time 61× While Boosting Accuracy in Diffusion LLMs

The article analyzes a recent Duke University paper that reveals a "scratchpad" mechanism in diffusion large language models, proposes the DPad method to prune redundant suffix tokens before decoding, and demonstrates up to 61.4× faster inference with unchanged or even improved accuracy across multiple benchmarks.

DPadInference Accelerationdiffusion LLM

0 likes · 10 min read

How DPad Cuts Inference Time 61× While Boosting Accuracy in Diffusion LLMs

AI Frontier Lectures

Apr 24, 2025 · Artificial Intelligence

How d1 Boosts Reasoning in Diffusion LLMs with Reinforcement Learning

Researchers from UCLA and Meta AI introduce d1, a two‑stage post‑training framework that combines supervised fine‑tuning and a novel diffu‑GRPO reinforcement‑learning algorithm to enable efficient reasoning in masked diffusion large language models, achieving state‑of‑the‑art performance on multiple math and logic benchmarks.

AIReinforcement learningd1

0 likes · 9 min read

How d1 Boosts Reasoning in Diffusion LLMs with Reinforcement Learning