llm-unlearning

Awesome Machine Unlearning (A Survey of Machine Unlearning)

A resource repository for machine unlearning in large language models

[NeurIPS 2025] The one-stop repository for LLM unlearning

[ACL 2024] Code and data for "Machine Unlearning of Pre-trained Large Language Models"

[NeurIPS25] Official repo for "Simplicity Prevails: Rethinking Negative Preference Optimization for LLM Unlearning"

[NeurIPS 2024] Large Language Model Unlearning via Embedding-Corrupted Prompts

ICLR 2026: Implementation of our unlearning method "Partial Model Collapse" introduced in: "Model Collapse Is Not a Bug but a Feature in Machine Unlearning for LLMs".

FALCON: Fine-grained Activation Manipulation for LLM Unlearning (NeurIPS 2025)

[EMNLP 2026 (Oral)] Measuring the Depth of LLM Unlearning via Activation Patching

The LLM Unlearning repository is an open-source project dedicated to the concept of unlearning in Large Language Models (LLMs). It aims to address concerns about data privacy and ethical AI by exploring and implementing unlearning techniques that allow models to forget unwanted or sensitive data. This ensures that AI models comply with privacy.

ICLR 2025 Oral: Implementation of the probabilistic evaluation framework introduced in the paper: "A Probabilistic Perspective on Unlearning and Alignment for Large Language Models" accepted at ICLR 2025 (Oral).

Usenix Security’25: Measuring Sample-level Unlearning Completeness

The project is for LLM unlearning, trustworthy AI.

AAAI-25: On Effects of Steering Latent Representation for Large Language Model Unlearning

[TMLR 2026] Improving LLM Unlearning Robustness via Random Perturbations

Hands-on machine unlearning course for Responsible AI: Make ML models and LLMs forget data (GDPR right to be forgotten) and prove it. NumPy notebooks, classical ML to LLMs.

Code for the paper "Forget What Comes Next: LLM Unlearning via Statistical Independence"

A curated list of papers, benchmarks, and resources on machine unlearning for multimodal large language models (MLLMs) and vision-language models (VLMs).

A continuously-updated catalog of machine unlearning papers with live citation counts.