Causal Interventions in Language Models

Follow this topic Ask about this topic

This field asks which internal states, features, heads, neurons, or context elements actually influence a language model’s decisions, rather than merely correlate with them. Researchers use counterfactual prompt and evidence edits, activation patching and ablation, causal tracing, sparse-autoencoder features, and inference-time steering or weight edits to test mechanisms and control behavior. Applications include checking whether chain-of-thought and retrieved evidence drive answers, and modifying refusal, hallucination, bias, tool use, or reasoning while preserving unrelated capabilities. Results show that readable representations are often not causally used or easily steered; effects vary by layer, task, model, and context, while redundancy and collateral effects make reliable, selective interventions a continuing challenge.

Latest papers

489 papers83 papers in the last four weeks, up 22% on the four weeks before. 0.5% of all new papers.

Hidden Error Awareness in Chain-of-Thought Reasoning: The Signal Is Diagnostic, Not Causal
May 10, 2026Aojie Yuan, Zhiyuan Julian Su, Haiyue Zhang +2LLM Interpretability Causal Interventions in Language Models
University of Southern California, Los Angeles, CA, USA.
Decomposing and Steering Functional Metacognition in Large Language Models
May 9, 2026Yanshi Li, Xueru Bai, Shuman Liu +2LLM Evaluation LLM Interpretability
Shopee Singapore, Singapore
The Echo Amplifies the Knowledge: Somatic Marker Analogues in Language Models via Emotion Vector Re-Injection
May 9, 2026Jared GloverMemory-Augmented Language Models Causal Interventions in Language Models
CapSen Robotics
A Single Neuron Is Sufficient to Bypass Safety Alignment in Large Language Models
May 8, 2026Hamid Kazemi, Atoosa Chegini, Maria SafiLLM Interpretability Adversarial Attacks on LLMs
Apple · Work done during an internship at Apple. Present affiliation: University of Maryland, College Park.
How Much Do Circuits Tell Us? Measuring the Consistency and Specificity of Language Model Circuits
May 8, 2026Michael Li, Nishant SubramaniCircuit Discovery Mechanistic Interpretability
Northeastern University · Carnegie Mellon University – Language Technologies Institute
Tool Calling is Linearly Readable and Steerable in Language Models
May 8, 2026Zekun Wu, Ze Wang, Seonglae Cho +4AI Agent Reliability Tool-Augmented Language Model Agents
University College London · Holistic AI · Imperial College London
Where's the Plan? Locating Latent Planning in Language Models with Lightweight Mechanistic Interventions
May 8, 2026Nicole Ma, Nick RuiLLM Planning Mechanistic Interpretability
Stanford University.
Inference Time Causal Probing in LLMs
May 8, 2026Sadegh Khorasani, Saber Salehkaleybar, Negar Kiyavash +1Causal Interventions in Language Models Activation Steering
School of Computer and Communication Sciences, EPFL„ Lausanne, Switzerland · 2Leiden Institute of Advanced Computer Science (LIACS), Leiden University, , Leiden, The Netherlands · College of Management of Technology, EPFL, Lausanne, Switzerland
Mathematical Reasoning via Intervention-Based Time-Series Causal Discovery Using LLMs as Concept Mastery Simulators
May 8, 2026Tsuyoshi OkitaMathematical Reasoning Causal Inference
Kyushu Institute of Technology
Rethinking Dense Sequential Chains: Reasoning Language Models Can Extract Answers from Sparse, Order-Shuffling Chain-of-Thoughts
May 8, 2026Yi-Chang Chen, Feng-Ting Liao, Da-shan Shiu +1CoT Reasoning Causal Interventions in Language Models
MediaTek Research · Artificial Intelligence Center of Research Excellence, National Taiwan University
Instruction Tuning Changes How Upstream State Conditions Late Readout: A Cross-Patching Diagnostic
May 8, 2026Yifan ZhouTransformer Interpretability Instruction Tuning
University of California, Los Angeles
The Convergence Gap: Instruction-Tuned Language Models Stabilize Later in the Forward Pass
May 8, 2026Yifan ZhouLLM Interpretability Next-Token Prediction
University of California, Los Angeles
Crafting Reversible SFT Behaviors in Large Language Models
May 7, 2026Yuping Lin, Pengfei He, Yue Xing +5Supervised Fine-Tuning Language Model Steering
Michigan State University · Hippocratic AI · University of Georgia
Patch-Effect Graph Kernels for LLM Interpretability
May 7, 2026Ruben Fernandez-Boullon, David N. OlivieriGraph Representation Learning Mechanistic Interpretability
Department of Computer Science, University of Vigo, Ourense 32004, Spain
Don't Lose Focus: Activation Steering via Key-Orthogonal Projections
May 7, 2026Haoyan Luo, Mateo Espinosa Zarlenga, Mateja JamnikLanguage Model Steering Self-Attention
University of Cambridge · University of Oxford
Towards Steering without Sacrifice: Principled Training of Steering Vectors for Prompt-only Interventions
May 7, 2026Yuntai Bao, Qinfeng Li, Xinyan Yu +6Language Model Steering Causal Interventions in Language Models
Zhejiang University · Innovation and Management Center, School of Software Technology (Ningbo), Zhejiang University · Ant Group
Causal Probing for Internal Visual Representations in Multimodal Large Language Models
May 7, 2026Zehao Deng, Tianjie Ju, Zheng Wu +5Multimodal Large Language Models VLM Interpretability
School of Computer Science, Shanghai Jiao Tong University · Ant Group
Manifold Steering Reveals the Shared Geometry of Neural Network Representation and Behavior
May 6, 2026Daniel Wurgaft, Can Rager, Matthew Kowal +13Neural Representation Geometry Manifold Learning
Stanford University · University College London · Northeastern University +2
Automatically Finding and Validating Unexpected Side-Effects of Interventions on Language Models
May 6, 2026Quintin Pope, Ajay Hayagreeve Balaji, Jacques Thibodeau +1Language Model Generation Evaluation LLM Auditing
Oregon State University · Independent Researcher
ReplaySCM: A Benchmark for Executable Causal Mechanism Induction from Interventions
May 5, 2026Serafim BatzoglouLLM Evaluation Structural Causal Models
Steer Like the LLM: Activation Steering that Mimics Prompting
May 5, 2026Geert Heyman, Frederik VandeputteLanguage Model Steering LLM Prompting
Nokia Bell Labs, Belgium.
Where Paths Split: Localized, Calibrated Control of Moral Reasoning in Large Language Models
May 5, 2026Chenchen Yuan, Zheyu Zhang, Gjergji KasneciLLM Alignment Moral Reasoning in Language Models
School of Computation, Information and Technology, Technical University of Munich · School of Social Sciences and Technology, Technical University of Munich · Munich Center for Machine Learning (MCML)
Pairwise matrices for sparse autoencoders: single-feature inspection mislabels causal axes
May 4, 2026Michael A. Riegler, Birk Sebastian Frostelid Torpmann-HagenSparse Autoencoders Mechanistic Interpretability
SimulaMet, Norway · Simula, Norway
Neuron-Anchored Rule Extraction for Large Language Models via Contrastive Hierarchical Ablation
May 4, 2026Francesco Sovrano, Gabriele Dominici, Marc LangheinrichLLM Interpretability Mechanistic Interpretability
Università della Svizzera italiana Lugano, Switzerland
Compared to What? Baselines and Metrics for Counterfactual Prompting
May 1, 2026Zihao Yang, Mosh Levy, Yoav Goldberg +1Counterfactual Evaluation Causal Interventions in Language Models
Northeastern University · Bar-Ilan University · Allen Institute for AI
Borrowed Geometry: Cross-Distribution Head-Importance Fingerprints of Frozen Pretrained Gemma 4 31B
May 1, 2026Abay BektursunCross-Modal Learning Attention Head Analysis
Independent research
Attention Is Where You Attack
Apr 30, 2026Aviral Srivastava, Sourav PandaSelf-Attention Adversarial Attacks
Amazon · The Pennsylvania State University
Are Tools All We Need? Unveiling the Tool-Use Tax in LLM Agents
Apr 30, 2026Kaituo Zhang, Zhen Xiong, Mingyu Zhong +4Tool-Augmented Language Model Agents Agentic Reasoning
University of Houston · University of Southern California · 3New York University +2
Minimal, Local, Causal Explanations for Jailbreak Success in Large Language Models
Apr 30, 2026Shubham Kumar, Narendra AhujaLLM Interpretability Causal Reasoning in Language Models
University of Illinois Urbana-Champaign
DPN-LE: Dual Personality Neuron Localization and Editing for Large Language Models
Apr 30, 2026Lifan Zheng, Xue Yang, Jiawei Chen +6LLM Interpretability Personality Modeling in Language Models
Southeast University · Shanghai Jiao Tong University · East China Normal University +6

May 10, 2026J/K move · Enter open · S save

Hidden Error Awareness in Chain-of-Thought Reasoning: The Signal Is Diagnostic, Not Causal

Aojie Yuan, Zhiyuan Julian Su, Haiyue Zhang, Yi Nian+1

University of Southern California, Los Angeles, CA, USA.

LLM Interpretability Causal Interventions in Language Models Reasoning Trace Analysis

Open paper AskPDF arXiv

Chain-of-thought (CoT) prompting assumes that generated reasoning reflects a model's internal computation. We show this assumption is wrong in a specific, measurable way: models internally detect their own reasoning errors but outwardly express confidence in them. A linear probe on hidden states predicts trace correctness with 0.95 AUROC -- from the very first reasoning step (0.79) -- while verbalized confidence for wrong traces is 4.55/5, nearly identical to correct ones (4.87/5). A text-surface classifier achieves only 0.59 on the same data, confirming a 0.20-point gap invisible in the generated text. This hidden error awareness holds across three model families (Qwen, Llama, Phi), 1.5B-72B parameters, and RL-trained reasoning models (DeepSeek-R1, 0.852 AUROC). The natural question is whether this signal can fix the errors it detects. It cannot. Four interventions -- activation steering, probe-guided best-of-N, self-correction, and activation patching -- all fail; patching destroys output coherence entirely. The signal is diagnostic, not causal: a readout of computation quality, not a lever to redirect it. This delineates a boundary for mechanistic interpretability: error representations during reasoning are fundamentally different from the factual knowledge representations that prior work has successfully edited.