Skip to content

@OPTML-Group OPTML Group

Change the repository type filter

All

    Repositories list

    36 repositories

    • OPTML-Group.github.io

      Public
      SCSS
      •2•1•0•0•Updated Oct 27, 2025Oct 27, 2025

    • Unlearn-Backdoor

      Public
      Python
      •
      GNU General Public License v3.0
      •0•0•1•0•Updated Oct 21, 2025Oct 21, 2025

    • Unlearn-FullStack

      Public
      Python
      •
      MIT License
      •0•2•0•0•Updated Oct 12, 2025Oct 12, 2025

    • Unlearn-R2MU

      Public
      Reasoning Model Unlearning: Forgetting Traces, Not Just Answers, While Preserving Reasoning Skills
      Python
      •0•1•0•0•Updated Oct 9, 2025Oct 9, 2025

    • Unlearn-Simple

      Public
      [NeurIPS25] Official repo for "Simplicity Prevails: Rethinking Negative Preference Optimization for LLM Unlearning"
      data-privacylanguage-modelmachine-unlearningtrustworthy-machine-learningtrustworthy-ailarge-language-modelsllm-unlearning
      Python
      •
      MIT License
      •9•34•1•1•Updated Oct 3, 2025Oct 3, 2025

    • Unlearn-Smooth

      Public
      [ICML25] Official repo for "Towards LLM Unlearning Resilient to Relearning Attacks: A Sharpness-Aware Minimization Perspective and Beyond"
      Python
      •
      MIT License
      •1•14•0•0•Updated Sep 27, 2025Sep 27, 2025

    • Unlearn-Trace

      Public
      Unlearning Isn't Invisible: Detecting Unlearning Traces in LLMs from Model Outputs
      machine-learningmachine-unlearningtrustworthy-aillmlarge-language-model
      Python
      •
      MIT License
      •1•22•0•0•Updated Jul 5, 2025Jul 5, 2025

    • CyclicReflex

      Public
      "CyclicReflex: Improving Large Reasoning Models via Cyclical Reflection Token Scheduling" by Chongyu Fan, Yihua Zhang, Jinghan Jia, Alfred Hero, Sijia Liu
      Python
      •
      MIT License
      •0•4•1•0•Updated Jun 22, 2025Jun 22, 2025

    • VLM-Safety-Unlearn

      Public
      Safety Mirage: How Spurious Correlations Undermine VLM Safety Fine-tuning
      machine-unlearningtrustworthy-machine-learningvision-language-modelmultimodal-large-language-modelssafety-alignment
      Python
      •
      MIT License
      •0•12•1•0•Updated Jun 17, 2025Jun 17, 2025

    • Unlearn-ILU

      Public
      Python
      •
      MIT License
      •0•4•0•0•Updated Jun 15, 2025Jun 15, 2025

    • EPiC

      Public
      Python
      •0•3•0•0•Updated Jun 11, 2025Jun 11, 2025

    • Unlearn-Saliency

      Public
      [ICLR24 (Spotlight)] "SalUn: Empowering Machine Unlearning via Gradient-based Weight Saliency in Both Image Classification and Generation" by Chongyu Fan*, Jiancheng Liu*, Yihua Zhang, Eric Wong, Dennis Wei, Sijia Liu
      generative-modeldata-privacydiffusionmachine-unlearningforgettingdiffusion-modelsdata-deletionmembership-inference-attackunlearningmembership-inference
      Python
      •
      MIT License
      •28•137•4•0•Updated May 27, 2025May 27, 2025

    • Unlearn-WorstCase

      Public
      [ECCV24] "Challenging Forgets: Unveiling the Worst-Case Forget Sets in Machine Unlearning" by Chongyu Fan*, Jiancheng Liu*, Alfred Hero, Sijia Liu
      evaluationdata-privacyevaluation-frameworkmachine-unlearningforgettingdata-deletionunlearningdata-removal
      Python
      •
      MIT License
      •2•23•2•0•Updated May 27, 2025May 27, 2025

    • MU-Coreset

      Public
      [COLM2025]"LLM Unlearning Reveals a Stronger-Than-Expected Coreset Effect in Current Benchmarks"
      Python
      •1•1•0•0•Updated Apr 22, 2025Apr 22, 2025

    • Diffusion-MU-Attack

      Public
      The official implementation of ECCV'24 paper "To Generate or Not? Safety-Driven Unlearned Diffusion Models Are Still Easy To Generate Unsafe Images ... For Now". This work introduces one fast and effective attack method to evaluate the harmful-content generation ability of safety-driven unlearned diffusion models.
      evaluation-frameworkrobustnessadversarial-attacksunlearningstable-diffusionattack-unlearned-diffusion-model
      Python
      •
      MIT License
      •4•86•4•0•Updated Feb 28, 2025Feb 28, 2025

    • WAGLE

      Public
      Official repo for NeurIPS'24 paper "WAGLE: Strategic Weight Attribution for Effective and Modular Unlearning in Large Language Models"
      Python
      •
      MIT License
      •3•15•1•0•Updated Dec 16, 2024Dec 16, 2024

    • UnlearnCanvas

      Public
      [NeurIPS 2024 D&B Track] UnlearnCanvas: A Stylized Image Dataset to Benchmark Machine Unlearning for Diffusion Models by Yihua Zhang, Chongyu Fan, Yimeng Zhang, Yuguang Yao, Jinghan Jia, Jiancheng Liu, Gaoyuan Zhang, Gaowen Liu, Ramana Kompella, Xiaoming Liu, Sijia Liu
      Python
      •2•77•14•0•Updated Nov 11, 2024Nov 11, 2024

    • AdvUnlearn

      Public
      Official implementation of NeurIPS'24 paper "Defensive Unlearning with Adversarial Training for Robust Concept Erasure in Diffusion Models". This work adversarially unlearns the text encoder to enhance the robustness of unlearned DMs against adversarial prompt attacks and achieves a better balance between unlearning performance and image generation
      robust-optimizationadversarial-machine-learningunlearningstable-diffusionunlearned-diffusion-model
      Jupyter Notebook
      •
      Creative Commons Attribution 4.0 International
      •2•49•1•0•Updated Nov 4, 2024Nov 4, 2024

    • DeepZero

      Public
      [ICLR'24] "DeepZero: Scaling up Zeroth-Order Optimization for Deep Model Training" by Aochuan Chen*, Yimeng Zhang*, Jinghan Jia, James Diffenderfer, Jiancheng Liu, Konstantinos Parasyris, Yihua Zhang, Zheng Zhang, Bhavya Kailkhura, Sijia Liu
      blackbox-optimizationzeroth-order-optimizationefficient-deep-learningai4science
      Python
      •
      MIT License
      •9•66•2•0•Updated Oct 9, 2024Oct 9, 2024

    • SOUL

      Public
      Official repo for EMNLP'24 paper "SOUL: Unlocking the Power of Second-Order Optimization for LLM Unlearning"
      Python
      •
      MIT License
      •6•28•1•0•Updated Oct 1, 2024Oct 1, 2024

    • QF-Attack

      Public
      [CVPR23W] "A Pilot Study of Query-Free Adversarial Attack against Stable Diffusion" by Haomin Zhuang, Yihua Zhang and Sijia Liu
      Python
      •2•26•0•0•Updated Aug 27, 2024Aug 27, 2024

    • BiBadDiff

      Public
      "From Trojan Horses to Castle Walls: Unveiling Bilateral Backdoor Effects in Diffusion Models" by Zhuoshi Pan*, Yuguang Yao*, Gaowen Liu, Bingquan Shen, H. Vicky Zhao, Ramana Rao Kompella, Sijia Liu
      Python
      •2•7•1•0•Updated Mar 25, 2024Mar 25, 2024

    • BackdoorMSPC

      Public
      [ICLR2024]"Backdoor Secrets Unveiled: Identifying Backdoor Data with Optimized Scaled Prediction Consistency" by Soumyadeep Pal, Yuguang Yao, Ren Wang, Bingquan Shen, Sijia Liu
      Python
      •0•3•1•0•Updated Mar 14, 2024Mar 14, 2024

    • Unlearn-Sparse

      Public
      [NeurIPS23 (Spotlight)] "Model Sparsity Can Simplify Machine Unlearning" by Jinghan Jia*, Jiancheng Liu*, Parikshit Ram, Yuguang Yao, Gaowen Liu, Yang Liu, Pranay Sharma, Sijia Liu
      data-privacymachine-unlearningforgettingdata-deletionmembership-inference-attackunlearningmembership-inferencedata-removal
      Python
      •
      MIT License
      •11•81•3•0•Updated Mar 12, 2024Mar 12, 2024

    • .github

      Public
      0•0•0•0•Updated Feb 11, 2024Feb 11, 2024

    • BLO-Toolbox

      Public
      1•1•0•0•Updated Dec 18, 2023Dec 18, 2023

    • DP4TL

      Public
      [NeurIPS2023] "Selectivity Drives Productivity: Efficient Dataset Pruning for Enhanced Transfer Learning" by Yihua Zhang*, Yimeng Zhang*, Aochuan Chen*, Jinghan Jia, Jiancheng Liu, Gaowen Liu, Mingyi Hong, Shiyu Chang, Sijia Liu
      Python
      •2•14•0•0•Updated Oct 12, 2023Oct 12, 2023

    • RED-adv

      Public
      [WACV25] "Can Adversarial Examples Be Parsed to Reveal Victim Model Information?" by Yuguang Yao*, Jiancheng Liu*, Yifan Gong*, Xiaoming Liu, Yanzhi Wang, Xue Lin, Sijia Liu
      Python
      •0•7•0•0•Updated Oct 5, 2023Oct 5, 2023

    • CLAW-SAT

      Public
      [SANER 2023] CLAWSAT: Towards Both Robust and Accurate Code Models.
      Python
      •
      MIT License
      •1•6•0•0•Updated Oct 5, 2023Oct 5, 2023

    • ILM-VP

      Public
      [CVPR23] "Understanding and Improving Visual Prompting: A Label-Mapping Perspective" by Aochuan Chen, Yuguang Yao, Pin-Yu Chen, Yihua Zhang, and Sijia Liu
      Python
      •
      MIT License
      •13•53•0•0•Updated Sep 17, 2023Sep 17, 2023