Repositories list

• Causal-Forcing

  Public

  thu-ml/Causal-Forcing’s past year of commit activity
  Official codebase for "Causal Forcing: Autoregressive Diffusion Distillation Done Right for High-Quality Real-Time Interactive Video Generation"

  diffusiondistillationvideo-generation+ 14world-modelsdiffusion-modelsconsistency-modelstext-to-videoautoregressive-modelsgenerative-aiworld-model+ 7

  Python

  •

  Apache License 2.0

  •12•338•5•0•Updated Feb 11, 2026Feb 11, 2026

• CausalForcing.github.io

  Public

  thu-ml/CausalForcing.github.io’s past year of commit activity
  JavaScript

  •0•0•0•0•Updated Feb 7, 2026Feb 7, 2026

• RDT2

  Public

  thu-ml/RDT2’s past year of commit activity
  Official code of RDT 2

  Python

  •

  Apache License 2.0

  •41•705•13•1•Updated Feb 7, 2026Feb 7, 2026

• ultraimage.github.io

  Public

  thu-ml/ultraimage.github.io’s past year of commit activity
  JavaScript

  •0•0•0•0•Updated Feb 6, 2026Feb 6, 2026

• DiT-Extrapolation

  Public

  thu-ml/DiT-Extrapolation’s past year of commit activity
  Official implementation for "RIFLEx: A Free Lunch for Length Extrapolation in Video Diffusion Transformers" (ICML 2025) and UltraViCo (ICLR 2026)

  generative-modelropediffusion+ 9extrapolationditvideo-generationposition-embeddingdiffusion-modelslong-video-generationdiffusion-transformer+ 2

  Python

  •

  Apache License 2.0

  •74•783•23•0•Updated Feb 2, 2026Feb 2, 2026

• TurboDiffusion

  Public

  thu-ml/TurboDiffusion’s past year of commit activity
  TurboDiffusion: 100–200× Acceleration for Video Diffusion Models

  rcmdistillationvideo-generation+ 7mlsystemdiffusion-modelsinference-accelerationai-infraconsistency-modelsageattentionsparse-linear-attention

  Python

  •

  Apache License 2.0

  •231•3.3k•59•7•Updated Jan 29, 2026Jan 29, 2026

• i-DODE

  Public

  thu-ml/i-DODE’s past year of commit activity
  Official code for "Improved Techniques for Maximum Likelihood Estimation for Diffusion ODEs" (ICML 2023)

  Python

  •

  Apache License 2.0

  •1•19•1•0•Updated Jan 27, 2026Jan 27, 2026

• embodied-data-toolkit

  Public

  thu-ml/embodied-data-toolkit’s past year of commit activity
  A toolkit for processing raw embodied data into standardized formats and converting between embodied dataset schemas.

  data-processingembodied-aiformat-converter-tool

  Python

  •2•15•0•0•Updated Jan 21, 2026Jan 21, 2026

• RoboticsDiffusionTransformer

  Public

  thu-ml/RoboticsDiffusionTransformer’s past year of commit activity
  RDT-1B: a Diffusion Foundation Model for Bimanual Manipulation

  Python

  •

  MIT License

  •150•1.6k•37•3•Updated Jan 21, 2026Jan 21, 2026

• SLA

  Public

  thu-ml/SLA’s past year of commit activity
  SLA: Beyond Sparsity in Diffusion Transformers via Fine-Tunable Sparse–Linear Attention

  transformervideo-generationmlsys+ 7inference-accelerationai-infralinear-attentionsparse-attentiondiffusion-transformertrain-accelerationsparse-linear-attention

  Python

  •

  Apache License 2.0

  •15•264•7•0•Updated Jan 17, 2026Jan 17, 2026

• SageAttention

  Public
  [ICLR2025, ICML2025, NeurIPS2025 Spotlight] Quantized Attention achieves speedup of 2-5x compared to FlashAttention, without losing end-to-end metrics across la…

  cudatritonattention+ 9vitquantizationvideo-generationmlsysinference-accelerationefficient-attentionllm+ 2

  Cuda

  •

  Apache License 2.0

  •346•3.2k•152•17•Updated Jan 17, 2026Jan 17, 2026

• MLA-Trust

  Public
  A toolbox for benchmarking Multimodal LLM Agents trustworthiness across truthfulness, controllability, safety and privacy dimensions through 34 interactive task…

  agentbenchmarkprivacy+ 6toolboxsafetymulti-modalcontrollabilitytrustworthy-aitruthfulness

  Python

  •

  MIT License

  •4•63•2•0•Updated Jan 9, 2026Jan 9, 2026

• Motus

  Public

  thu-ml/Motus’s past year of commit activity
  Official code of Motus: A Unified Latent Action World Model

  roboticsdiffusion-modelvideo-generation+ 4robotic-manipulationworld-modelunidiffuservision-language-action-model

  Python

  •

  Apache License 2.0

  •27•768•20•0•Updated Jan 5, 2026Jan 5, 2026

• SpargeAttn

  Public
  [ICML2025] SpargeAttention: A training-free sparse attention that accelerates any model inference.

  attentionvitquantization+ 8video-generationmlsysinference-accelerationai-infravision-transformersparse-attentionllm+ 1

  Cuda

  •

  Apache License 2.0

  •84•938•57•3•Updated Dec 31, 2025Dec 31, 2025

• vidar-robotwin

  Public

  thu-ml/vidar-robotwin’s past year of commit activity
  robotwin evaluation code for vidar.

  Python

  •

  MIT License

  •0•4•1•0•Updated Dec 22, 2025Dec 22, 2025

• vidar

  Public
  Official repo for vidar and vidarc: video foundation model for robotics.

  Python

  •1•37•1•0•Updated Dec 22, 2025Dec 22, 2025

• UltraViCo.github.io

  Public

  thu-ml/UltraViCo.github.io’s past year of commit activity
  Project page for "UltraViCo"

  JavaScript

  •0•0•0•0•Updated Dec 3, 2025Dec 3, 2025

• tianshou

  Public

  thu-ml/tianshou’s past year of commit activity
  An elegant PyTorch deep reinforcement learning library.

  pytorchdqnpolicy-gradient+ 16rlcqlatariddpgimitation-learningsacdrl+ 9

  Python

  •

  MIT License

  •1.3k•10k•132•2•Updated Dec 1, 2025Dec 1, 2025

• UniCardio

  Public

  thu-ml/UniCardio’s past year of commit activity
  Versatile Cardiovascular Signal Generation with a Unified Diffusion Transformer

  Python

  •

  MIT License

  •3•50•2•0•Updated Oct 9, 2025Oct 9, 2025

• FrameBridge

  Public
  FrameBridge: Improving Image-to-Video Generation with Bridge Models

  Python

  •0•3•0•0•Updated Oct 2, 2025Oct 2, 2025

• DiffusionBridge

  Public

  thu-ml/DiffusionBridge’s past year of commit activity
  Official codebase for "Diffusion Bridge Implicit Models" (ICLR 2025) and "Consistency Diffusion Bridge Models" (NeurIPS 2024)

  Python

  •10•104•6•0•Updated Sep 9, 2025Sep 9, 2025

• SparseDM

  Public

  thu-ml/SparseDM’s past year of commit activity
  Python

  •1•1•0•0•Updated Aug 24, 2025Aug 24, 2025

• EmbodiedActiveDefense

  Public
  Official implementation for "Reinforced Embodied Active Defense: Exploiting Adaptive Interaction for Robust Visual Perception in Adversarial 3D Environments" (T…

  Python

  •

  MIT License

  •0•8•0•0•Updated Jul 26, 2025Jul 26, 2025

• oddefense

  Public
  Jupyter Notebook

  •

  MIT License

  •2•5•0•0•Updated Jul 16, 2025Jul 16, 2025

• Adaptive-Sparse-Trainer

  Public
  Official implementation for "Pruning Large Language Models with Semi-Structural Adaptive Sparse Training" (AAAI 2025)

  Python

  •

  Apache License 2.0

  •2•18•0•0•Updated Jul 1, 2025Jul 1, 2025

• MMTrustEval

  Public

  thu-ml/MMTrustEval’s past year of commit activity
  A toolbox for benchmarking trustworthiness of multimodal large language models (MultiTrust, NeurIPS 2024 Track Datasets and Benchmarks)

  benchmarkprivacytoolbox+ 9safetymulti-modalfairnessrobustnessclaudegpt-4trustworthy-ai+ 2

  Python

  •

  Creative Commons Attribution Share Alike 4.0 International

  •12•174•1•0•Updated Jun 27, 2025Jun 27, 2025

• TetraJet-MXFP4Training

  Public

  thu-ml/TetraJet-MXFP4Training’s past year of commit activity
  Pytorch implementation of "Oscillation-Reduced MXFP4 Training for Vision Transformers" on DeiT Model Pre-training

  Python

  •

  Apache License 2.0

  •4•36•0•0•Updated Jun 20, 2025Jun 20, 2025

• DDO

  Public
  Python

  •

  Other

  •0•0•0•0•Updated Jun 16, 2025Jun 16, 2025

• GFT

  Public
  Python

  •

  MIT License

  •0•51•5•0•Updated Jun 13, 2025Jun 13, 2025

• cond-image-leakage

  Public
  Official implementation for "Identifying and Solving Conditional Image Leakage in Image-to-Video Diffusion Model" (NeurIPS 2024)

  Apache License 2.0

  •22•258•8•0•Updated Apr 25, 2025Apr 25, 2025