Aug 2026· Proceedings of the 32nd ACM SIGKDD Conference on Knowledge Discovery and Data Mining V.2· 0 citations· 5 references
TL;DR
This tutorial provides a comprehensive, end-to-end view of LLM interpretability, transitioning from microscopic neural analysis to macroscopic application and deployment, and explores how these interpretability paradigms scale and inspire the design of frontier architectures, agentic systems, and thinking models.
Abstract
The rapid evolution of Large Language Models (LLMs) has brought unprecedented capabilities across reasoning, coding, and multimodal tasks. However, as performance scales, their opaque ''black-box'' nature raises a critical challenge: How can we trace the origins of emergent intelligence, and more importantly, how can we leverage these internal mechanisms to guide model optimization? This tutorial provides a comprehensive, end-to-end view of LLM interpretability, transitioning from microscopic neural analysis to macroscopic application and deployment. It is systematically organized into five core sections: i) Unlocking the Black Box: We begin with the evolution of LLM interpretability and highlight recent breakthroughs from leading research teams. ii) Methodology: We present a rigorous overview of foundational theories (e.g., mathematical framework for transformer, biological mechanisms in LLMs) and essential methods (e.g., path patching, logit lens, and neuron description). iii) Anatomy of LLMs: Using advanced techniques to decode internal semantic features, neural circuits, and complex behaviors, we interpret how models perform reasoning, factual recall, and in-context learning. iv) Applications: We show how to transfer interpretability insights into actionable improvements across the LLM pipeline, including interpretability-guided data synthesis (data value scoring, corpus filtering, and activation-based data diagnosis). We also present Pinpoint Training and Steering for precise capability gains, and Pinpoint Quantization for extreme low-bit compression with minimal capability loss. v) Advanced Topics: We conclude by exploring how these interpretability paradigms scale and inspire the design of frontier architectures, agentic systems, and thinking models. In this tutorial, researchers and engineers will gain the theoretical frameworks and practical engineering toolkits needed to understand, steer, and efficiently deploy LLMs in real-world production environments.
Large language models (LLMs) have demonstrated remarkable capabilities across a wide range of tasks, yet their internal mechanisms remain largely opaque, making it difficult to understand, predict, or control their behavior. As LLMs are increasingly deployed in high-stakes settings, this lack of transparency raises serious concerns about reliability and safety. Mechanistic interpretability (MI) has emerged as a promising approach to address this challenge, seeking to reverse-engineer the internal computations of LLMs into human-understandable mechanisms, i.e., an approximate high-level algorithm that the LLM implements with a subset of its components (a circuit) to complete a certain language task or exhibit a certain behavior. This tutorial provides a comprehensive and up-to-date overview of LLM mechanism discovery, validation, and editing. We begin by introducing foundational concepts, including features, components, computational graphs, and circuits, along with key notation. We then examine mechanism discovery through four methodological families: causal mediation, attribution, sparse decomposition, and optimization-based approaches. Next, we turn to mechanism validation, covering methods for verifying proposed mechanisms and emerging standards for rigorous evaluation. Building on these foundations, we survey mechanistic editing techniques that leverage MI insights to modify behavior at varying granularity, from fine-grained representation-level steering to coarser circuit-level interventions. Lastly, we outline open challenges and future research directions, including scalability of interpretability methods, evaluation benchmarks for mechanistic circuits, and the integration of interpretability with training-time objectives, aiming to inspire continued progress in understanding and governing large language models.
Yinhan He, Wendy Zheng, Tianyi Zhao et al.· Proceedings of the 32nd ACM...· 0 citations
Interpretability is often treated as a tax on capability: language models are trained as opaque systems, then explained after the fact, with methods whose reliability is difficult to establish. In this work, we challenge this premise. Rather than reverse-engineering a model, we make interpretability a constraint of the training pipeline, optimized alongside the language modeling objective. Across three orders of magnitude of compute, on both autoregressive and diffusion language models, interpretability scales with capability rather than against it. Surprisingly, model representations become more disentangled and aligned with human-understandable concepts with scale. We instantiate the training-time recipe with Steerling-8B, a diffusion language model with a causal attention mask. For any group of generated tokens, Steerling-8B attributes the output to relevant input tokens, human-understandable concepts, and training data. This enables closed-loop intervention: diagnose an output through its concept or feature attribution, retrieve similar training data, and correct the behavior through concept steering without retraining. Steerling-8B remains competitive with open peer models trained on substantially 2-16x more compute, suggesting a different scaling paradigm: interpretability can be designed into training, and it improves with scale.
Guide Labs Team, Andreas Madsen, A. Ismail et al.· 3 citations· ⚡1
It is indicated that RL-trained models not only demonstrate greater awareness of their learned behaviors and stronger generalizability to novel, structurally similar tasks than SFT models but also often exhibit weak alignment between their reasoning traces and final outputs, an effect most pronounced in GRPO-trained models.
Pratham Singla, Shivank Garg, Ayush Singh et al.· Annual Meeting of the Associ...· 0 citations
The black box nature of deep neural networks poses a significant challenge for the deployment of transparent and trustworthy artificial intelligence (AI) systems. With the growing presence of AI in society, it becomes increasingly important to develop methods that can explain and interpret the decisions made by these systems. To address this, mechanistic interpretability (MI) emerged as a promising and distinctive research program within the broader field of explainable artificial intelligence (XAI). MI studies the inner computations of neural networks and translates them into human-understandable algorithms. It encompasses reverse-engineering techniques aimed at uncovering the computational algorithms implemented by neural networks. In this paper, we present a comprehensive survey of mechanistic interpretability, synthesizing a rapidly growing and fragmented body of work into a single, structured reference. We propose a unified taxonomy of MI approaches and provide a detailed analysis of key techniques, illustrated with concrete examples and pseudo-code. We contextualize MI within the broader interpretability landscape, comparing its goals, methods, and insights to other strands of XAI. Additionally, we trace the development of MI as a research area, highlighting its conceptual roots and the accelerating pace of recent work. We argue that MI has the potential to support a more scientific understanding of machine learning systems – treating models not only as tools for solving tasks, but also as systems to be studied and understood. We intend this survey to serve as an entry point for new researchers to the field of mechanistic interpretability.
This article offers a comprehensive overview of mechanistic interpretability, an emerging field that seeks to reverse-engineer the internal algorithms of modern neural networks. While traditional explainable AI methods often stop at surface-level input-output correlations, this approach directly addresses the opaque"black box"nature of machine learning models, which is essential for ensuring safety and auditability in high-stakes deployments. The paper provides a detailed examination of Transformer circuit analysis, exploring how internal components like the residual stream, attention mechanisms, and induction heads drive complex tasks and in-context learning. It subsequently tackles the core challenge of superposition and polysemanticity, demonstrating how tools like Sparse Autoencoders (SAEs) and transcoders can decompose tangled network activations into distinct, human-interpretable features. Furthermore, the paper explores methods for actively controlling and modifying model behavior through steering vectors and causal interventions. Finally, it connects these mechanistic insights with neurosymbolic AI frameworks designed to translate neural representations into explicit, executable logical rules.