arXiv:2607. 19317v1 Announce Type: new Abstract: Circuit analysis can support not only model explanation but also downstream interventions such as pruning, editing, steering, and selective fine-tuning.
arXiv:2605.09129v2 Announce Type: replace Abstract: Circuit study aims to explain how language models (LMs) implement a task by localizing and interpreting a circuit: a computational subgraph respons...
arXiv:2606. 08236v1 Announce Type: cross Abstract: As large language models are increasingly deployed in high-stakes settings, there is a growing need for tools that audit not only model outputs but also the internal computations that produce them.
The paper introduces an enhanced end‑to‑end circuit analysis framework built on Gemini 2.5 Pro, targeting engineering education. It addresses two key failure modes—circuit‑recognition hallucinations and reasoning‑process hallucinations—by adding a YOLO detector for source polarity re‑identification and an ngspice verification loop for iterative refinement. The resulting pipeline achieves 97.59 % accuracy on 83 undergraduate problems, markedly outperforming the baseline Gemini model and demonstrating significant gains across varied diagram styles and textbooks.
S^3martCirc is a self‑supervised framework that jointly discovers and interprets neural circuits in large language models, rather than treating circuit discovery and functional interpretation as separate stages. It abstracts node behavior into two general computational roles that generalize across tasks and introduces a quantitative metric for assigning these roles, enabling simultaneous identification of importance and function. Experiments demonstrate that S^3martCirc outperforms existing methods in circuit discovery.
arXiv:2606. 24026v1 Announce Type: new Abstract: Mechanistic interpretability has made substantial progress in automatically localizing circuits, but explaining what localized components do remains labor-intensive and difficult to standardize.
arXiv:2606. 16939v1 Announce Type: cross Abstract: A prominent research direction in mechanistic interpretability is learning sparse circuits over LLM components to reveal how they jointly produce model behavior.
The paper investigates whether structural differences in circuits discovered by circuit discovery methods reflect distinct mechanisms. By varying input-token frequency while keeping the task constant, the authors find that although circuits appear specialized by frequency structurally, functional and representational analyses reveal no reliable differences, a phenomenon they call phantom specialization. Across multiple models and tasks, structurally distinct circuits implement the same computation, with core shared subgraphs recovering most of the performance and interchangeable internal representations confirmed by causal interventions.
arXiv:2603. 09161v2 Announce Type: replace-cross Abstract: Learning effective netlist representations is fundamentally constrained by the scarcity of labeled datasets, as real designs are protected by Intellectual Property (IP) and costly to annotate.
arXiv:2504. 03711v2 Announce Type: replace-cross Abstract: Artificial intelligence (AI)-driven electronic design automation (EDA) techniques have been extensively explored for VLSI circuit design applications.
arXiv:2609.15668v1 Announce Type: cross Abstract: Through pre-training on extensive text and image datasets, current multi-modal large language models (MLLMs) achieve strong performance on general ta...
The paper investigates mechanistic interpretability, focusing on how automated circuit discovery is evaluated. It shows that the commonly used faithfulness objective can favor circuits that reproduce a model’s behavior poorly, creating an objective-level recovery gap. Experiments on four human-reference tasks and InterpBench reveal that many discovery methods misrank candidate circuits, and that restoring excluded signals can correct most of these misrankings without altering the circuits’ behavior.
arXiv:2608. 09374v1 Announce Type: new Abstract: Electrical circuit analysis requires more than recognizing components in an image.