arXiv Machine Learning June 30, 2026 By Barsat Khadka

MechRL: Reinforcement Learning Agents Perform Circuit Discovery for Mechanistic Interpretability

Read the original on arXiv Machine Learning →

arXiv:2605. 26343v2 Announce Type: replace Abstract: Mechanistic interpretability seeks to explain a model's behaviour by finding its circuit: the sparse subgraph of the model's computation that is causally responsible for it.

Summary generated by The Flow from the publisher's feed. The full article lives at arXiv Machine Learning.

arXiv AI

Jul 22

Circuit Claims Depend on What Is Extracted and How It Is Compared

arXiv:2607. 18921v1 Announce Type: cross Abstract: Circuit extraction identifies a small set of model components whose presence preserves a target behavior under ablation, and the resulting circuit is often read as the mechanism behind that behavior.

By Yang Sheng, Jie Fu

llms reinforcement-learning efficiency benchmarks

arXiv AI

Jun 9

Mechanistic origins of catastrophic forgetting: why RL preserves circuits better than SFT?

arXiv:2605. 28860v2 Announce Type: replace-cross Abstract: Fine-tuning large language models (LLMs) frequently induces catastrophic forgetting of prior capabilities.

By Jeanmely Rojas Nunez, Viraj Sawant, Nathan Allen, Nomgondalai Amgalanbaatar, Yannis Zongo, Vasu Sharma, Maheep Chaudhary

llms reinforcement-learning fine-tuning

arXiv Machine Learning

Aug 11

Can Graph Learning Learn Circuits?

arXiv:2608. 08536v1 Announce Type: new Abstract: Circuit localization is a mechanistic interpretability task whose goal is to identify a sparse subgraph of a transformer's computation graph sufficient to reproduce a particular behavior.

By Chester Tan, Moritz Lampert, Courtney Maynard, Ankit Ramakrishnan, Tina Eliassi-Rad, Ingo Scholtes

llms benchmarks safety

arXiv AI

Jun 11

ATLAS: Active Theory Learning for Automated Science

arXiv:2606. 12386v1 Announce Type: cross Abstract: Advancing scientific understanding through mechanistic modeling requires posing the right experimental questions to yield maximally informative data.

By No\'emi \'Eltet\H{o}, Nathaniel D. Daw, Kimberly L. Stachenfeld, Kevin J. Miller

agents reinforcement-learning

arXiv Machine Learning

Jun 25

FactorLibrary: From Polynomials to Circuits via Recursive Subgoals

arXiv:2606. 25394v1 Announce Type: new Abstract: Finding minimal arithmetic circuits for polynomials over finite fields is a combinatorially hard problem central to algebraic complexity theory.

By Rohan Pandey, Michael Ruofan Zeng, Weikun K. Zhang, Kaijie Jin, Naomi Morato, Archit Ganapule, Bhaumik Mehta, Jarod Alper

agents reinforcement-learning

arXiv AI

Jul 28

SeekJudge: A Practical Reward Framework for Reinforcement Learning in Computer-Use Agents

arXiv:2607. 23263v1 Announce Type: new Abstract: Deciding whether a trajectory actually fulfills its instruction governs how we measure computer-use agents on long-horizon graphical-user-interface tasks and how we train them with reinforcement learning.

By Yang Wan, Zhenhao Zhang, Jierui Wang, Linchao Zhu

agents reinforcement-learning efficiency

Related stories

Circuit Claims Depend on What Is Extracted and How It Is Compared

Mechanistic origins of catastrophic forgetting: why RL preserves circuits better than SFT?

Can Graph Learning Learn Circuits?

ATLAS: Active Theory Learning for Automated Science

FactorLibrary: From Polynomials to Circuits via Recursive Subgoals

SeekJudge: A Practical Reward Framework for Reinforcement Learning in Computer-Use Agents