A curated collection of resources focused on the Mechanistic Interpretability (MI) of Large Multimodal Models (LMMs). This repository aggregates surveys, blog posts, and research papers that explore how LMMs represent, transform, and align multimodal information internally.
☆223Mar 4, 2026Updated 7 months ago
Alternatives and similar repositories for Awesome-LMMs-Mechanistic-Interpretability
Users that are interested in Awesome-LMMs-Mechanistic-Interpretability are comparing it to the libraries listed below. We may earn a commission when you buy through links labeled 'Ad' on this page.
Sorting:
- awesome SAE papers☆81May 24, 2025Updated last year
- ☆37Jun 13, 2025Updated last year
- ☆15Jan 20, 2026Updated 8 months ago
- ☆273Nov 22, 2024Updated last year
- awesome papers in LLM interpretability☆628Aug 20, 2025Updated last year
- Deploy on Railway without the complexity - Free Credits Offer • AdConnect your repo and Railway handles the rest with instant previews. Quickly provision container image services, databases, and storage volumes.
- A curated list of resources for activation engineering☆140Oct 2, 2025Updated last year
- Training Sparse Autoencoders on Language Models☆1,563Updated this week
- [ICCV 2025] Auto Interpretation Pipeline and many other functionalities for Multimodal SAE Analysis.☆203Sep 26, 2025Updated last year
- [ACL 2025] "CoT-UQ: Improving Response-wise Uncertainty Quantification in LLMs with Chain-of-Thought"☆17Apr 3, 2025Updated last year
- Sparse autoencoders for vision☆67Sep 25, 2026Updated 3 weeks ago
- [ICML'25] Our study systematically investigates massive values in LLMs' attention mechanisms. First, we observe massive values are concen…☆87Jun 20, 2025Updated last year
- [ICCV25 Oral] Token Activation Map to Visually Explain Multimodal LLMs☆191Dec 14, 2025Updated 10 months ago
- ☆87Nov 5, 2024Updated last year
- A carefully curated collection of high-quality libraries, projects, tutorials, research papers, and other essential resources focused on …☆166Updated this week
- Managed Kubernetes at scale on DigitalOcean • AdDigitalOcean Kubernetes includes the control plane, bandwidth allowance, container registry, automatic updates, and more for free.
- [NeurIPS 2025] Official Implementation of paper "Sherlock: Self-Correcting Reasoning in Vision-Language Models"☆31Jun 4, 2026Updated 4 months ago
- [ICLR 2026] Official implementation of the paper "Map the Flow: Revealing Hidden Pathways of Information in VideoLLMs"☆26Mar 3, 2026Updated 7 months ago
- A library of visualization tools for the interpretability and hallucination analysis of large vision-language models (LVLMs).☆43May 22, 2025Updated last year
- [NeurIPS'26] In this work, we investigate how Large Language Models (LLMs) adapt their internal representations when encountering inputs …☆29Oct 10, 2026Updated last week
- Less is More: Mitigating Multimodal Hallucination from an EOS Decision Perspective (ACL 2024)☆59Oct 28, 2024Updated last year
- The Github repo for our survey paper: "Locate, Steer, and Improve: A Practical Survey of Actionable Mechanistic Interpretability in Large…☆160Apr 15, 2026Updated 6 months ago
- A library for mechanistic interpretability of GPT-style language models☆3,975Updated this week
- A framework that allows you to apply Sparse AutoEncoder on any models☆53Jul 11, 2025Updated last year
- A versatile toolkit for applying Logit Lens to modern large language models (LLMs). Currently supports Llama-3.1-8B and Qwen-2.5-7B, enab…☆178Aug 14, 2025Updated last year
- Deploy on Railway without the complexity - Free Credits Offer • AdConnect your repo and Railway handles the rest with instant previews. Quickly provision container image services, databases, and storage volumes.
- FeatureAlignment = Alignment + Mechanistic Interpretability☆35Mar 8, 2025Updated last year
- [CVPR 2025] Mitigating Object Hallucinations in Large Vision-Language Models with Assembly of Global and Local Attention☆71Jul 16, 2024Updated 2 years ago
- [ACL 2025] Data and Code for Paper VLSBench: Unveiling Visual Leakage in Multimodal Safety☆62Jul 21, 2025Updated last year
- ViT Prisma is a mechanistic interpretability library for Vision and Video Transformers (ViTs).☆394Oct 7, 2026Updated last week
- A repository for awesome resources in mechanistic interpretability☆17Jan 18, 2023Updated 3 years ago
- code for EMNLP 2024 paper: Neuron-Level Knowledge Attribution in Large Language Models☆52Nov 17, 2024Updated last year
- ☆16Nov 14, 2025Updated 11 months ago
- The official repo for "Where do Large Vision-Language Models Look at when Answering Questions?"☆74Jan 7, 2026Updated 9 months ago
- Code for my NeurIPS 2024 ATTRIB paper titled "Attribution Patching Outperforms Automated Circuit Discovery"☆51May 31, 2024Updated 2 years ago
- Proton VPN Special Offer - Get 70% off • AdSpecial partner offer. Trusted by over 100 million users worldwide. Tested, Approved and Recommended by Experts.
- [EMNLP 2025 Demo] Extracting internal representations from vision-language models. Beta version.☆130Apr 25, 2026Updated 5 months ago
- NeurIPS 2026: Mechanistic Interpretability toolkit for Vision-Language-Action models☆29Oct 1, 2026Updated 2 weeks ago
- [ICLR '25] Official Pytorch implementation of "Interpreting and Editing Vision-Language Representations to Mitigate Hallucinations"☆105Nov 30, 2025Updated 10 months ago
- Mechanistic Interpretability Visualizations using React☆368Apr 30, 2026Updated 5 months ago
- This repository collects all relevant resources about interpretability in LLMs☆410Nov 1, 2024Updated last year
- MemEye: A Visual-Centric Evaluation Framework for Multimodal Agent Memory☆56May 17, 2026Updated 5 months ago
- ☆213Nov 17, 2024Updated last year