mechanistic interpretability

ReDeEP: Detecting Hallucination in Retrieval-Augmented Generation via Mechanistic Interpretability ICLR 2025 Residual Stream Analysis with Multi-Layer SAEs ICLR 2025 RomanLens: The Role Of Latent Romanization In Multilinguality In LLMs ACL 2025 Route Sparse Autoencoder to Interpret Large Language Models EMNLP 2025 SCIURus: Shared Circuits for Interpretable Uncertainty Representations in Language Models NAACL 2025 Separating Tongue from Thought: Activation Patching Reveals Language-Agnostic Concept Representations in Transformers ACL 2025 Short-circuiting Shortcuts: Mechanistic Investigation of Shortcuts in Text Classification CoNLL 2025 Sparse Activation Editing for Reliable Instruction Following in Narratives EMNLP 2025 Stochastic Chameleons: Irrelevant Context Hallucinations Reveal Class-Based (Mis)Generalization in LLMs ACL 2025 The Same but Different: Structural Similarities and Differences in Multilingual Language Modeling ICLR 2025 Time Course MechInterp: Analyzing the Evolution of Components and Knowledge in Large Language Models ACL 2025 TinySQL: A Progressive Text-to-SQL Dataset for Mechanistic Interpretability Research EMNLP 2025 Toward Efficient Sparse Autoencoder-Guided Steering for Improved In-Context Learning in Large Language Models EMNLP 2025 Towards a Unified and Verified Understanding of Group-Operation Networks ICLR 2025 Towards Global-level Mechanistic Interpretability: A Perspective of Modular Circuits of Large Language Models ICML 2025 Towards Universality: Studying Mechanistic Similarity Across Language Model Architectures ICLR 2025 Transformers Struggle to Learn to Search ICLR 2025 Understanding and Controlling Repetition Neurons and Induction Heads in In-Context Learning IJCNLP 2025 Understanding How Value Neurons Shape the Generation of Specified Values in LLMs EMNLP 2025 Understanding the Repeat Curse in Large Language Models from a Feature Perspective ACL 2025 Understanding Verbatim Memorization in LLMs Through Circuit Discovery ACL 2025 Unraveling the Influence of Training Data and Internal Structures in Large Language Models for Enhanced Explainability (Student Abstract) AAAI 2025 Validating Mechanistic Interpretations: An Axiomatic Approach ICML 2025 What Do VLMs NOTICE? A Mechanistic Interpretability Pipeline for Gaussian-Noise-free Text-Image Corruption and Evaluation NAACL 2025 Where Confabulation Lives: Latent Feature Discovery in LLMs EMNLP 2025