AM
Aaron Mueller
cs.CLcs.AIcs.LGComputational LinguisticsComputational Modelingcs.CVcs.CYdemographic inferenceFirst and Second Language AcquisitionLanguage
On Valency
published · living versionsW_s9uxv46y·v1 · currentpublished
Interventional Training Reduces Causal Bypass in In-Context Introspection
with Zilu Tang, Jongin Kim, Kerem Sahin, David Atkinson +1
1 version
Preprints & journals
49 papers in the corpus · 2019–2026The Curse of Multiple Mediators: Hidden Interaction Effects in Activation Patching2606.27510v1 · Sankaran Vaidyanathan, David Arbour, Aaron Mueller et al.2026 · 0 citationsarXiv
Do Language Models Track Entities Across State Changes?2605.30233v1 · Zilu Tang, Qiao Zhao, Gabriel Franco et al.2026 · 0 citationsarXiv
Algorithm that gets 'under the hood' of AI models could effectively steer their responses.42056297 · Mueller, Aaron2026 · 0 citationsNature. 2026;653(8113):35-36
Latent Agents: A Post-Training Procedure for Internalized Multi-Agent Debate2604.24881v1 · John Seon Keun Yi, Aaron Mueller, Dokyun Lee2026 · 0 citationsarXiv
Pitfalls in Evaluating Interpretability Agents2603.20101v1 · Tal Haklay, Nikhil Prakash, Sana Pandey et al.2026 · 0 citationsarXiv
BabyLM Turns 4 and Goes Multilingual: Call for Papers for the 2026 BabyLM Workshop2602.20092v2 · Leshem Choshen, Ryan Cotterell, Mustafa Omer Gul et al.2026 · 0 citationsarXiv
Causality is Key for Interpretability Claims to Generalise2602.16698v2 · Shruti Joshi, Aaron Mueller, David Klindt et al.2026 · 0 citationsarXiv
Measuring Mechanistic Independence: Can Bias Be Removed Without Erasing Demographics?2512.20796v1 · Zhengyang Shan, Aaron Mueller2025 · 0 citationsarXiv
From Isolation to Entanglement: When Do Interpretability Methods Identify and Disentangle Known Concepts?2512.15134v2 · Aaron Mueller, Andrew Lee, Shruti Joshi et al.2025 · 0 citationsarXiv
BabyVLM-V2: Toward Developmentally Grounded Pretraining and Benchmarking of Vision Foundation Models2512.10932v2 · Shengao Wang, Wenqi Wang, Zecheng Wang et al.2025 · 0 citationsarXiv
Findings of the BlackboxNLP 2025 Shared Task: Localizing Circuits and Causal Variables in Language Models2511.18409v1 · Dana Arad, Yonatan Belinkov, Hanjie Chen et al.2025 · 1 citationarXiv
Crosscoding Through Time: Tracking Emergence & Consolidation Of Linguistic Representations Throughout LLM Pretraining2509.05291v2 · Deniz Bayazit, Aaron Mueller, Antoine Bosselut2025 · 0 citationsarXiv
CRISP: Persistent Concept Unlearning via Sparse Autoencoders2508.13650v3 · Tomer Ashuach, Dana Arad, Aaron Mueller et al.2025 · 0 citationsarXiv
SAEs Are Good for Steering -- If You Select the Right Features2505.20063v2 · Dana Arad, Aaron Mueller, Yonatan Belinkov2025 · 2 citationsarXiv
Improving the OOD Performance of Closed-Source LLMs on NLI Through Strategic Data Selection2505.20209v2 · Joe Stacey, Lisa Alazraki, Aran Ubhi et al.2025 · 0 citationsarXiv
MIB: A Mechanistic Interpretability Benchmark2504.13151v2 · Aaron Mueller, Atticus Geiger, Sarah Wiegreffe et al.2025 · 1 citationarXiv
Findings of the BabyLM Challenge: Sample-Efficient Pretraining on Developmentally Plausible Corpora2504.08165v1 · Alex Warstadt, Aaron Mueller, Leshem Choshen et al.2025 · 74 citations2023. In Proceedings of the BabyLM Challenge at the 27th Conference on Computational Natural Language Learning, pages 1-34, Singapore. Association for Computational Linguistics
BabyLM Turns 3: Call for papers for the 2025 BabyLM workshop2502.10645v2 · Lucas Charpentier, Leshem Choshen, Ryan Cotterell et al.2025 · 0 citationsarXiv
Position-aware Automatic Circuit Discovery2502.04577v1 · Tal Haklay, Hadas Orgad, David Bau et al.2025 · 2 citationsarXiv
Disjoint Processing Mechanisms of Hierarchical and Linear Grammars in Large Language Models2501.08618v1 · Aruna Sankaranarayanan, Dylan Hadfield-Menell, Aaron Mueller2025 · 0 citationsarXiv
Large Language Models Share Representations of Latent Grammatical Concepts Across Typologically Diverse Languages2501.06346v2 · Jannik Brinkmann, Chris Wendler, Christian Bartelt et al.2025 · 5 citationsarXiv
Findings of the Second BabyLM Challenge: Sample-Efficient Pretraining on Developmentally Plausible Corpora2412.05149v1 · Michael Y. Hu, Aaron Mueller, Candace Ross et al.2024 · 3 citationsarXiv
Incremental Sentence Processing Mechanisms in Autoregressive Transformer Language Models2412.05353v1 · Michael Hanna, Aaron Mueller2024 · 3 citationsarXiv
Characterizing the Role of Similarity in the Property Inferences of Language Models2410.22590v2 · Juan Diego Rodriguez, Aaron Mueller, Kanishka Misra2024 · 1 citationarXiv
Arithmetic Without Algorithms: Language Models Solve Math With a Bag of Heuristics2410.21272v2 · Yaniv Nikankin, Anja Reusch, Aaron Mueller et al.2024 · 4 citationsarXiv
The Quest for the Right Mediator: Surveying Mechanistic Interpretability Through the Lens of Causal Mediation Analysis2408.01416v3 · Aaron Mueller, Jannik Brinkmann, Millicent Li et al.2024 · 2 citationsarXiv
NNsight and NDIF: Democratizing Access to Open-Weight Foundation Model Internals2407.14561v4 · Jaden Fiotto-Kaufman, Alexander R. Loftus, Eric Todd et al.2024 · 4 citationsarXiv
Bigger is not always better: The importance of human-scale language modeling for psycholinguisticsrfwgd_v2 · Ethan Gotlieb Wilcox, Michael Hu, Aaron Mueller et al.2024 · 9 citationsPsyArXiv
Missed Causes and Ambiguous Effects: Counterfactuals Pose Challenges for Interpreting Neural Networks2407.04690v1 · Aaron Mueller2024 · 0 citationsarXiv
[Call for Papers] The 2nd BabyLM Challenge: Sample-efficient pretraining on a developmentally plausible corpus2404.06214v2 · Leshem Choshen, Ryan Cotterell, Michael Y. Hu et al.2024 · 2 citationsarXiv
Sparse Feature Circuits: Discovering and Editing Interpretable Causal Graphs in Language Models2403.19647v3 · Samuel Marks, Can Rager, Eric J. Michaud et al.2024 · 4 citationsInternational Conference on Learning Representations, 2025
In-context Learning Generalizes, But Not Always Robustly: The Case of Syntax2311.07811v2 · Aaron Mueller, Albert Webson, Jackson Petty et al.2023 · 8 citationsarXiv
Function Vectors in Large Language Models2310.15213v2 · Eric Todd, Millicent L. Li, Arnab Sen Sharma et al.2023 · 5 citationsarXiv
Meta-training with Demonstration Retrieval for Efficient Few-shot Learning2307.00119v1 · Aaron Mueller, Kanika Narang, Lambert Mathias et al.2023 · 0 citationsarXiv
Inverse Scaling: When Bigger Isn't Better2306.09479v2 · Ian R. McKenzie, Alexander Lyzhov, Michael Pieler et al.2023 · 29 citationsTransactions on Machine Learning Research (TMLR), 10/2023, https://openreview.net/forum?id=DwgRm72GQF
How to Plant Trees in Language Models: Data and Architectural Effects on the Emergence of Syntactic Inductive Biases2305.19905v1 · Aaron Mueller, Tal Linzen2023 · 11 citationsarXiv
Call for Papers -- The BabyLM Challenge: Sample-efficient pretraining on a developmentally plausible corpus2301.11796v1 · Alex Warstadt, Leshem Choshen, Aaron Mueller et al.2023 · 14 citationsarXiv
Language model acceptability judgements are not always robust to context2212.08979v1 · Koustuv Sinha, Jon Gauthier, Aaron Mueller et al.2022 · 6 citationsarXiv
Causal Analysis of Syntactic Agreement Neurons in Multilingual Language Models2210.14328v1 · Aaron Mueller, Yu Xia, Tal Linzen2022 · 5 citationsarXiv
What Do NLP Researchers Believe? Results of the NLP Community Metasurvey2208.12852v1 · Julian Michael, Ari Holtzman, Alicia Parrish et al.2022 · 26 citationsarXiv
Label Semantic Aware Pre-training for Few-shot Text Classification2204.07128v2 · Aaron Mueller, Jason Krone, Salvatore Romeo et al.2022 · 39 citationsarXiv
Coloring the Blank Slate: Pre-training Imparts a Hierarchical Inductive Bias to Sequence-to-sequence Models2203.09397v1 · Aaron Mueller, Robert Frank, Tal Linzen et al.2022 · 17 citationsarXiv
Causal Analysis of Syntactic Agreement Mechanisms in Neural Language Models2106.06087v3 · Matthew Finlayson, Aaron Mueller, Sebastian Gehrmann et al.2021 · 44 citationsarXiv
Fine-tuning Encoders for Improved Monolingual and Zero-shot Polylingual Neural Topic Modeling2104.05064v1 · Aaron Mueller, Mark Dredze2021 · 11 citationsarXiv
Demographic Representation and Collective Storytelling in the Me Too Twitter Hashtag Activism Movement.35295189 · Mueller, Aaron, Wood-Doughty, Zach, Amir, Silvio et al.2022 · 50 citationsProceedings of the ACM on human-computer interaction. 2021;5(CSCW1)
Decoding Methods for Neural Narrative Generation2010.07375v2 · Alexandra DeLucia, Aaron Mueller, Xiang Lisa Li et al.2020 · 13 citationsarXiv
Cross-Linguistic Syntactic Evaluation of Word Prediction Models2005.00187v2 · Aaron Mueller, Garrett Nicolai, Panayiota Petrou-Zeniou et al.2020 · 42 citationsarXiv
Modeling Color Terminology Across Thousands of Languages1910.01531v1 · Arya D. McCarthy, Winston Wu, Aaron Mueller et al.2019 · 8 citationsarXiv
Quantity doesn't buy quality syntax with neural language models1909.00111v1 · Marten van Schijndel, Aaron Mueller, Tal Linzen2019 · 68 citationsarXiv
Career total: 66 works. 49 are in this corpus.
Profile built from the corpus for this byline.
Author records are still filling in while the Hub is in alpha. If this is your page, you'll be able to claim it soon. Spot a mistake? Tell us.