DENG Lab @ SJTU ORG

Discrete Diffusion Forcing (D2F): dLLMs Can Do Faster-Than-AR Inference

No description.

[CVPR 2026] Mantis: A Versatile Vision-Language-Action Model with Disentangled Visual Foresight

No description.

The official implementation of World-Language-Action Model for Unified World Modeling, Language Reasoning, and Action Synthesis

A placeholder repository for the World-Language-Action Model from SJTU's Deng Lab, an AI research project combining environment world modeling, language reasoning, and action synthesis, with code and weights not yet released.

Flexible and Pluggable Serving Engine for Diffusion LLMs

SIFT: Grounding LLM Reasoning in Contexts via Stickers

No description.

Official implementation for "LOVECon: Text-driven Training-free Long Video Editing with ControlNet"

No description.

No description.

LoPA: Scaling dLLM Inference via Lookahead Parallel Decoding

LightningRL: Breaking the Accuracy–Parallelism Trade-off of Block-wise dLLMs via Reinforcement Learning

[Findings of EMNLP 2024] AdaMoE: Token-Adaptive Routing with Null Experts for Mixture-of-Experts Language Models

Code for orthogonal neural operator

No description.

Multi-Block Diffusion Language Models

No description.

No description.

No description.

High Performance LLM Inference Operator Library

GPU Monitor

None