activation patching

Beyond Transcription: Mechanistic Interpretability in ASR AAAI 2026 Bridging Linguistic Structure and Mechanistic Interpretability for Conceptual Interpretation in Language Models CoNLL 2026 Can you map it to English? The Role of Cross-Lingual Alignment in the Multilingual Performance of LLMs EACL 2026 Fine-Grained Analysis of Shared Syntactic Mechanisms in Language Models ACL 2026 Mechanistic Interpretability of Animacy Effects on Structure Choice in GPT-2 CoNLL 2026 Spectra: A Mechanistic Interpretability Library for Vision-Language Models ACL 2026 When Meanings Meet: Investigating the Emergence and Quality of Shared Concept Spaces during Multilingual Language Model Training EACL 2026 When Truth Is Overridden: Uncovering the Internal Origins of Sycophancy in Large Language Models AAAI 2026 Answer, Assemble, Ace: Understanding How LMs Answer Multiple Choice Questions ICLR 2025 Discursive Circuits: How Do Language Models Understand Discourse Relations? EMNLP 2025 Dissecting Persona-Driven Reasoning in Language Models via Activation Patching EMNLP 2025 Elucidating Mechanisms of Demographic Bias in LLMs for Healthcare EMNLP 2025 From Reasoning to Answer: Empirical, Attention-Based and Mechanistic Insights into Distilled DeepSeek R1 Models EMNLP 2025 Interpretability Analysis of Arithmetic In-Context Learning in Large Language Models EMNLP 2025 RomanLens: The Role Of Latent Romanization In Multilinguality In LLMs ACL 2025 Separating Tongue from Thought: Activation Patching Reveals Language-Agnostic Concept Representations in Transformers ACL 2025 SPIRIT: Patching Speech Language Models against Jailbreak Attacks EMNLP 2025 Towards Faithful Natural Language Explanations: A Study Using Activation Patching in Large Language Models EMNLP 2025 Why Do Some Inputs Break Low-Bit LLM Quantization? EMNLP 2025 Attribution Patching Outperforms Automated Circuit Discovery EMNLP 2024 Dissecting Fine-Tuning Unlearning in Large Language Models EMNLP 2024 Fine-Tuning Enhances Existing Mechanisms: A Case Study on Entity Tracking ICLR 2024 How Language Models Prioritize Contextual Grammatical Cues? EMNLP 2024 Is This the Subspace You Are Looking for? An Interpretability Illusion for Subspace Activation Patching ICLR 2024 LLMs to Replace Crowdsourcing For Parallel Data Creation? The Case of Text Detoxification EMNLP 2024