NeurIPS 2025 Workshop MechInterp Submissions

Towards Understanding Multimodal Fine-Tuning: A Case Study into Spatial Features
Lachin Naghashyar, Hunar Batra, Ashkan Khakzar, Philip Torr, Ronald Clark, Christian Schroeder de Witt, Constantin Venhoff
- Published: 30 Sept 2025, Last Modified: 30 Sept 2025
- Mech Interp Workshop (NeurIPS 2025) Poster
- Readers: Everyone
nnterp: A Standardized Interface for Mechanistic Interpretability of Transformers
Clément Dumas
- Published: 30 Sept 2025, Last Modified: 30 Sept 2025
- Mech Interp Workshop (NeurIPS 2025) Poster
- Readers: Everyone
Emergent Specialization: Rare Token Neurons in Language Models
Jing Liu, Yueheng Li, Haozheng Wang
- Published: 30 Sept 2025, Last Modified: 30 Sept 2025
- Mech Interp Workshop (NeurIPS 2025) Poster
- Readers: Everyone
TopKLoRA
Marek Masiak, Lukas Vierling, Constantin Venhoff, Nicola Cancedda, Christian Schroeder de Witt
- Published: 30 Sept 2025, Last Modified: 30 Sept 2025
- Mech Interp Workshop (NeurIPS 2025) Poster
- Readers: Everyone
Context Matters: Analyzing the Generalizability of Linear Probing and Steering Across Diverse Scenarios
Isha Agarwal, Saharsha Navani, Fazl Barez
- Published: 30 Sept 2025, Last Modified: 30 Sept 2025
- Mech Interp Workshop (NeurIPS 2025) Poster
- Readers: Everyone
High-order Component Attribution via Kolmogorov-Arnold Networks
Samy Mammeri, Christian Gagné
- Published: 30 Sept 2025, Last Modified: 30 Sept 2025
- Mech Interp Workshop (NeurIPS 2025) Poster
- Readers: Everyone
Small Vectors, Big Effects: A Mechanistic Study of RL-Induced Reasoning via Steering Vectors
Viacheslav Sinii, Nikita Balagansky, Yaroslav Aksenov, Vadim Kurochkin, Daniil Laptev, Alexey Gorbatovski, Boris Shaposhnikov, Daniil Gavrilov
- Published: 30 Sept 2025, Last Modified: 30 Sept 2025
- Mech Interp Workshop (NeurIPS 2025) Poster
- Readers: Everyone
Interpreting Vision Grounding in Vision-Language Models: A Case Study in Coordinate Prediction
Clement Neo, Yongsen Zheng, Kwok-Yan Lam, Luke Ong
- Published: 30 Sept 2025, Last Modified: 30 Sept 2025
- Mech Interp Workshop (NeurIPS 2025) Poster
- Readers: Everyone
Reverse Engineering a Stateful Reasoning circuit
Akshit Kumar, Dipti Sharma, Parameswari Krishnamurthy
- Published: 30 Sept 2025, Last Modified: 30 Sept 2025
- Mech Interp Workshop (NeurIPS 2025) Poster
- Readers: Everyone
Comparing Clinical and General LLMs on Knowledge Boundaries and Robustness
Xingmeng Zhao, Ke Yang, Anthony Rios
- Published: 30 Sept 2025, Last Modified: 30 Sept 2025
- Mech Interp Workshop (NeurIPS 2025) Poster
- Readers: Everyone
Chain-of-Thought Resampling for Interpreting LLM Decision-Making
Uzay Macar, Paul C. Bogdan, Senthooran Rajamanoharan, Neel Nanda
- Published: 30 Sept 2025, Last Modified: 30 Sept 2025
- Mech Interp Workshop (NeurIPS 2025) Poster
- Readers: Everyone
Discovering Transformer Circuits via a Hybrid Attribution and Pruning Framework
Hao Gu, Vibhas Nair, Amrithaa Ashok Kumar, Ryan Lagasse, Kevin Zhu, Sean O'Brien, Ashwinee Panda
- Published: 30 Sept 2025, Last Modified: 30 Sept 2025
- Mech Interp Workshop (NeurIPS 2025) Poster
- Readers: Everyone
Symbolic vs. Continuous Features in Transformers: A Digital Communication System's Explanation
Kan Deng
- Published: 30 Sept 2025, Last Modified: 30 Sept 2025
- Mech Interp Workshop (NeurIPS 2025) Poster
- Readers: Everyone
PII Jailbreaking in LLMs via Activation Steering Reveals Personal Information Leakage
Krishna Kanth Nakka, Xue Jiang, Dmitrii Usynin, Xuebing Zhou
- Published: 30 Sept 2025, Last Modified: 30 Sept 2025
- Mech Interp Workshop (NeurIPS 2025) Poster
- Readers: Everyone
Thinking Hard, Going Misaligned: Emergent Misalignment in LLMs
Hanqi Yan, Hainiu Xu, Yulan He
- Published: 30 Sept 2025, Last Modified: 30 Sept 2025
- Mech Interp Workshop (NeurIPS 2025) Poster
- Readers: Everyone
Learning Interpretable Features in Audio Latent Spaces via Sparse Autoencoders
Nathan Paek, Yongyi Zang, Qihui Yang, Randal Leistikow
- Published: 30 Sept 2025, Last Modified: 30 Sept 2025
- Mech Interp Workshop (NeurIPS 2025) Poster
- Readers: Everyone
Iterative Inference in a Chess-Playing Neural Network
Elias Sandmann, Sebastian Lapuschkin, Wojciech Samek
- Published: 30 Sept 2025, Last Modified: 30 Sept 2025
- Mech Interp Workshop (NeurIPS 2025) Poster
- Readers: Everyone
Activation Transport Operators
Andrzej Szablewski, Marek Masiak
- Published: 30 Sept 2025, Last Modified: 30 Sept 2025
- Mech Interp Workshop (NeurIPS 2025) Spotlight
- Readers: Everyone
Superposition in Mixture of Experts
Marmik Chaudhari, Jeremi Nuer, Rome Thorstenson
- Published: 30 Sept 2025, Last Modified: 30 Sept 2025
- Mech Interp Workshop (NeurIPS 2025) Poster
- Readers: Everyone
Death by a Thousand Directions: Exploring the Geometry of Harmfulness in LLMs through Subconcept Probing
Saleena Angeline Sartawita, McNair Shah, Adhitya Rajendra Kumar, Naitik Chheda, Will Cai, Kevin Zhu, Sean O'Brien, Vasu Sharma
- Published: 30 Sept 2025, Last Modified: 30 Sept 2025
- Mech Interp Workshop (NeurIPS 2025) Poster
- Readers: Everyone
Hallucination Reduction with CASAL: Contrastive Activation Steering for Amortized Learning
Wannan Yang, Xinchi Qiu, Lei Yu, Yuchen Zhang, Aobo Yang, Narine Kokhlikyan, Nicola Cancedda, Diego Garcia-Olano
- Published: 30 Sept 2025, Last Modified: 30 Sept 2025
- Mech Interp Workshop (NeurIPS 2025) Poster
- Readers: Everyone
Breaking the Mirror: Activation-Based Mitigation of Self-Preference in LLM Evaluators
Dani Roytburg, Matthew Nguyen, Matthew Bozoukov, Jou Barzdukas, Hongyu Fu, Narmeen Fatimah Oozeer
- Published: 30 Sept 2025, Last Modified: 30 Sept 2025
- Mech Interp Workshop (NeurIPS 2025) Poster
- Readers: Everyone
Analyzing and Improving Chain-of-Thought Monitorability Through Information Theory
Usman Anwar, Tim Bakker, Dana Kianfar, Cristina Pinneri, Christos Louizos
- Published: 30 Sept 2025, Last Modified: 02 Oct 2025
- Mech Interp Workshop (NeurIPS 2025) Poster
- Readers: Everyone
On the Geometry and Topology of Neural Circuits for Modular Addition
Gabriela Moisescu-Pareja, Gavin McCracken, Harley Wiltzer, Colin Daniels, Vincent Létourneau, Jonathan Love
- Published: 30 Sept 2025, Last Modified: 30 Sept 2025
- Mech Interp Workshop (NeurIPS 2025) Poster
- Readers: Everyone
Emerging Human-like Strategies for Semantic Memory Foraging in Large Language Models
Eric Lacosse, Mariana Duarte, Peter Todd, Daniel C McNamee
- Published: 30 Sept 2025, Last Modified: 30 Sept 2025
- Mech Interp Workshop (NeurIPS 2025) Poster
- Readers: Everyone