Proceedings of the 9th BlackboxNLP Workshop: Analyzing and Interpreting Neural Networks for NLP

Proceedings of the 9th BlackboxNLP Workshop: Analyzing and Interpreting Neural Networks for NLP Dana Arad editor Yonatan Belinkov editor Hanjie Chen editor Nils Feldhus editor Aaron Mueller editor Hosein Mohebbi editor Gabriele Sarti editor Martin Tutek editor 2026-10 Association for Computational Linguistics Budapest, Hungary text conference publication 979-8-89176-476-7 blackboxnlp-2026-1 https://aclanthology.org/2026.blackboxnlp-1.0/ import pandas as np: the numpy strikes back Haotian Yang author 2026-10 text Proceedings of the 9th BlackboxNLP Workshop: Analyzing and Interpreting Neural Networks for NLP Dana Arad editor Yonatan Belinkov editor Hanjie Chen editor Nils Feldhus editor Aaron Mueller editor Hosein Mohebbi editor Gabriele Sarti editor Martin Tutek editor Association for Computational Linguistics Budapest, Hungary conference publication 979-8-89176-476-7 yang-2026-import https://aclanthology.org/2026.blackboxnlp-1.1/ 2026-10 1 14 On The Effectiveness-Fluency Trade-Off In LLM Conditioning: A Systematic Study Iuri Macocco author Pau Rodriguez author Arno Blaas author Luca Zappella author Marco Baroni author Xavier Suau author 2026-10 text Proceedings of the 9th BlackboxNLP Workshop: Analyzing and Interpreting Neural Networks for NLP Dana Arad editor Yonatan Belinkov editor Hanjie Chen editor Nils Feldhus editor Aaron Mueller editor Hosein Mohebbi editor Gabriele Sarti editor Martin Tutek editor Association for Computational Linguistics Budapest, Hungary conference publication 979-8-89176-476-7 macocco-etal-2026-effectiveness https://aclanthology.org/2026.blackboxnlp-1.2/ 2026-10 15 33 Cultural Binding Heads in Language Models Avrile Floro author Luca Benedetto author 2026-10 text Proceedings of the 9th BlackboxNLP Workshop: Analyzing and Interpreting Neural Networks for NLP Dana Arad editor Yonatan Belinkov editor Hanjie Chen editor Nils Feldhus editor Aaron Mueller editor Hosein Mohebbi editor Gabriele Sarti editor Martin Tutek editor Association for Computational Linguistics Budapest, Hungary conference publication 979-8-89176-476-7 floro-benedetto-2026-cultural https://aclanthology.org/2026.blackboxnlp-1.3/ 2026-10 34 48 Lost in Specialization: Embedding Geometry and Retrieval Quality Across Specialized Corpora Cornelia Ferner author Stefan Wegenkittl author 2026-10 text Proceedings of the 9th BlackboxNLP Workshop: Analyzing and Interpreting Neural Networks for NLP Dana Arad editor Yonatan Belinkov editor Hanjie Chen editor Nils Feldhus editor Aaron Mueller editor Hosein Mohebbi editor Gabriele Sarti editor Martin Tutek editor Association for Computational Linguistics Budapest, Hungary conference publication 979-8-89176-476-7 ferner-wegenkittl-2026-lost https://aclanthology.org/2026.blackboxnlp-1.4/ 2026-10 49 57 Compact Language, Complex Model Shifts: How and Where Ambiguity and Underspecification Affect LLMs Michaela Regneri author Nina Scheller author Sören Laue author 2026-10 text Proceedings of the 9th BlackboxNLP Workshop: Analyzing and Interpreting Neural Networks for NLP Dana Arad editor Yonatan Belinkov editor Hanjie Chen editor Nils Feldhus editor Aaron Mueller editor Hosein Mohebbi editor Gabriele Sarti editor Martin Tutek editor Association for Computational Linguistics Budapest, Hungary conference publication 979-8-89176-476-7 regneri-etal-2026-compact https://aclanthology.org/2026.blackboxnlp-1.5/ 2026-10 58 75 How Much Do Modern Embeddings Reveal? Inversion Attacks on Commercial Embedding Models Kareem Elaasar author Cristina Lopes author 2026-10 text Proceedings of the 9th BlackboxNLP Workshop: Analyzing and Interpreting Neural Networks for NLP Dana Arad editor Yonatan Belinkov editor Hanjie Chen editor Nils Feldhus editor Aaron Mueller editor Hosein Mohebbi editor Gabriele Sarti editor Martin Tutek editor Association for Computational Linguistics Budapest, Hungary conference publication 979-8-89176-476-7 elaasar-lopes-2026-much https://aclanthology.org/2026.blackboxnlp-1.6/ 2026-10 76 89 Does the Greater-Than Circuit Generalize? A Reproducibility Study Across Models and Numerical Entailment Vinay Setty author 2026-10 text Proceedings of the 9th BlackboxNLP Workshop: Analyzing and Interpreting Neural Networks for NLP Dana Arad editor Yonatan Belinkov editor Hanjie Chen editor Nils Feldhus editor Aaron Mueller editor Hosein Mohebbi editor Gabriele Sarti editor Martin Tutek editor Association for Computational Linguistics Budapest, Hungary conference publication 979-8-89176-476-7 setty-2026-greater https://aclanthology.org/2026.blackboxnlp-1.7/ 2026-10 90 102 A circuit for predicting hierarchical structure in-context in Large Language Models Tankred Saanum author Can Demircan author Samuel J Gershman author Eric Schulz author 2026-10 text Proceedings of the 9th BlackboxNLP Workshop: Analyzing and Interpreting Neural Networks for NLP Dana Arad editor Yonatan Belinkov editor Hanjie Chen editor Nils Feldhus editor Aaron Mueller editor Hosein Mohebbi editor Gabriele Sarti editor Martin Tutek editor Association for Computational Linguistics Budapest, Hungary conference publication 979-8-89176-476-7 saanum-etal-2026-circuit https://aclanthology.org/2026.blackboxnlp-1.8/ 2026-10 103 121 Inside self-reflection: the geometry in multi-dimensional scoring of metacognitive structure Praveen Suthaharan author Philip Robert Corlett author Yuen-Siang Ang author 2026-10 text Proceedings of the 9th BlackboxNLP Workshop: Analyzing and Interpreting Neural Networks for NLP Dana Arad editor Yonatan Belinkov editor Hanjie Chen editor Nils Feldhus editor Aaron Mueller editor Hosein Mohebbi editor Gabriele Sarti editor Martin Tutek editor Association for Computational Linguistics Budapest, Hungary conference publication 979-8-89176-476-7 suthaharan-etal-2026-inside https://aclanthology.org/2026.blackboxnlp-1.9/ 2026-10 122 135 Data Attribution of Emergent Misalignment with Persona Features Clemens Vetter author David Kaczér author Lucie Flek author Florian Mai author 2026-10 text Proceedings of the 9th BlackboxNLP Workshop: Analyzing and Interpreting Neural Networks for NLP Dana Arad editor Yonatan Belinkov editor Hanjie Chen editor Nils Feldhus editor Aaron Mueller editor Hosein Mohebbi editor Gabriele Sarti editor Martin Tutek editor Association for Computational Linguistics Budapest, Hungary conference publication 979-8-89176-476-7 vetter-etal-2026-data https://aclanthology.org/2026.blackboxnlp-1.10/ 2026-10 136 165 Specialized Attention Heads in Early Word Reconstruction Gustaf Gren author Robert Östling author Murathan Kurfali author 2026-10 text Proceedings of the 9th BlackboxNLP Workshop: Analyzing and Interpreting Neural Networks for NLP Dana Arad editor Yonatan Belinkov editor Hanjie Chen editor Nils Feldhus editor Aaron Mueller editor Hosein Mohebbi editor Gabriele Sarti editor Martin Tutek editor Association for Computational Linguistics Budapest, Hungary conference publication 979-8-89176-476-7 gren-etal-2026-specialized https://aclanthology.org/2026.blackboxnlp-1.11/ 2026-10 166 177 From Plausible to Actionable: A Position on LLM Self-Explanations Elize Herrewijnen author Benedetta Muscato author Gizem Gezici author Fosca Giannotti author 2026-10 text Proceedings of the 9th BlackboxNLP Workshop: Analyzing and Interpreting Neural Networks for NLP Dana Arad editor Yonatan Belinkov editor Hanjie Chen editor Nils Feldhus editor Aaron Mueller editor Hosein Mohebbi editor Gabriele Sarti editor Martin Tutek editor Association for Computational Linguistics Budapest, Hungary conference publication 979-8-89176-476-7 herrewijnen-etal-2026-plausible https://aclanthology.org/2026.blackboxnlp-1.12/ 2026-10 178 188 How Far Do SAE Features Persist Across LLM Layers? An Empirical Study with an FFN Pruning Application Vaishakh Sreekanth Menon author Asif Ekbal author 2026-10 text Proceedings of the 9th BlackboxNLP Workshop: Analyzing and Interpreting Neural Networks for NLP Dana Arad editor Yonatan Belinkov editor Hanjie Chen editor Nils Feldhus editor Aaron Mueller editor Hosein Mohebbi editor Gabriele Sarti editor Martin Tutek editor Association for Computational Linguistics Budapest, Hungary conference publication 979-8-89176-476-7 menon-ekbal-2026-far https://aclanthology.org/2026.blackboxnlp-1.13/ 2026-10 189 200 Decodability Is Not Computation: Examining Probing Claims Kabir Singh Grewal author 2026-10 text Proceedings of the 9th BlackboxNLP Workshop: Analyzing and Interpreting Neural Networks for NLP Dana Arad editor Yonatan Belinkov editor Hanjie Chen editor Nils Feldhus editor Aaron Mueller editor Hosein Mohebbi editor Gabriele Sarti editor Martin Tutek editor Association for Computational Linguistics Budapest, Hungary conference publication 979-8-89176-476-7 grewal-2026-decodability https://aclanthology.org/2026.blackboxnlp-1.14/ 2026-10 201 219 Source Marking in Language Models: A Controlled Framework for Acquisition, Cue Sensitivity, and Reasoning Daniil Huryn author 2026-10 text Proceedings of the 9th BlackboxNLP Workshop: Analyzing and Interpreting Neural Networks for NLP Dana Arad editor Yonatan Belinkov editor Hanjie Chen editor Nils Feldhus editor Aaron Mueller editor Hosein Mohebbi editor Gabriele Sarti editor Martin Tutek editor Association for Computational Linguistics Budapest, Hungary conference publication 979-8-89176-476-7 huryn-2026-source https://aclanthology.org/2026.blackboxnlp-1.15/ 2026-10 220 231 Lost in Perception: Isolating Perceptual and Reasoning Failures in Multimodal Physics and Geometry Reasoning Raj Jaiswal author Sree Krishna Uppalapati author Dhruvkumar Patel author Ria Khatoniar author Tanuja Ganu author Rajiv Ratn Shah author 2026-10 text Proceedings of the 9th BlackboxNLP Workshop: Analyzing and Interpreting Neural Networks for NLP Dana Arad editor Yonatan Belinkov editor Hanjie Chen editor Nils Feldhus editor Aaron Mueller editor Hosein Mohebbi editor Gabriele Sarti editor Martin Tutek editor Association for Computational Linguistics Budapest, Hungary conference publication 979-8-89176-476-7 jaiswal-etal-2026-lost https://aclanthology.org/2026.blackboxnlp-1.16/ 2026-10 232 268 Probing for Knowledge Attribution in Large Language Models Ivo Brink author Alexander Boer author Dennis Ulmer author 2026-10 text Proceedings of the 9th BlackboxNLP Workshop: Analyzing and Interpreting Neural Networks for NLP Dana Arad editor Yonatan Belinkov editor Hanjie Chen editor Nils Feldhus editor Aaron Mueller editor Hosein Mohebbi editor Gabriele Sarti editor Martin Tutek editor Association for Computational Linguistics Budapest, Hungary conference publication 979-8-89176-476-7 brink-etal-2026-probing https://aclanthology.org/2026.blackboxnlp-1.17/ 2026-10 269 293 Induction in Both Directions: A Mechanistic Analysis of In-Context Learning in Masked Diffusion Language Models Andy Eduard Catruna author Emilian Radoi author 2026-10 text Proceedings of the 9th BlackboxNLP Workshop: Analyzing and Interpreting Neural Networks for NLP Dana Arad editor Yonatan Belinkov editor Hanjie Chen editor Nils Feldhus editor Aaron Mueller editor Hosein Mohebbi editor Gabriele Sarti editor Martin Tutek editor Association for Computational Linguistics Budapest, Hungary conference publication 979-8-89176-476-7 catruna-radoi-2026-induction https://aclanthology.org/2026.blackboxnlp-1.18/ 2026-10 294 303 What Does BERT Know about Linguistic Features: Probing Finnish Model for Text Difficulty Yiheng Wu author Jue Hou author Roman Yangarber author 2026-10 text Proceedings of the 9th BlackboxNLP Workshop: Analyzing and Interpreting Neural Networks for NLP Dana Arad editor Yonatan Belinkov editor Hanjie Chen editor Nils Feldhus editor Aaron Mueller editor Hosein Mohebbi editor Gabriele Sarti editor Martin Tutek editor Association for Computational Linguistics Budapest, Hungary conference publication 979-8-89176-476-7 wu-etal-2026-bert https://aclanthology.org/2026.blackboxnlp-1.19/ 2026-10 304 315 Where Animacy Lives in Large Language Models: Tracing the Circuits of the Animacy Concept Samuele Punzo author Giovanni Cinà author Sandro Pezzelle author 2026-10 text Proceedings of the 9th BlackboxNLP Workshop: Analyzing and Interpreting Neural Networks for NLP Dana Arad editor Yonatan Belinkov editor Hanjie Chen editor Nils Feldhus editor Aaron Mueller editor Hosein Mohebbi editor Gabriele Sarti editor Martin Tutek editor Association for Computational Linguistics Budapest, Hungary conference publication 979-8-89176-476-7 punzo-etal-2026-animacy https://aclanthology.org/2026.blackboxnlp-1.20/ 2026-10 316 348 Do Chess Explanations Reflect Model Decisions? Behavioral and Token-Level Tests of LLM Reasoning Faithfulness Angelina Parfenova author 2026-10 text Proceedings of the 9th BlackboxNLP Workshop: Analyzing and Interpreting Neural Networks for NLP Dana Arad editor Yonatan Belinkov editor Hanjie Chen editor Nils Feldhus editor Aaron Mueller editor Hosein Mohebbi editor Gabriele Sarti editor Martin Tutek editor Association for Computational Linguistics Budapest, Hungary conference publication 979-8-89176-476-7 parfenova-2026-chess https://aclanthology.org/2026.blackboxnlp-1.21/ 2026-10 349 367 Do Large Language Models appeal to the Senses? Narrative Embodiment in Stories told by Llama and Qwen Stefan Arnold author Rene Gröbner author 2026-10 text Proceedings of the 9th BlackboxNLP Workshop: Analyzing and Interpreting Neural Networks for NLP Dana Arad editor Yonatan Belinkov editor Hanjie Chen editor Nils Feldhus editor Aaron Mueller editor Hosein Mohebbi editor Gabriele Sarti editor Martin Tutek editor Association for Computational Linguistics Budapest, Hungary conference publication 979-8-89176-476-7 arnold-grobner-2026-large https://aclanthology.org/2026.blackboxnlp-1.22/ 2026-10 368 373 Reverse Engineering the Reversal Curse: What Entity Order in Pretraining Does and Does Not Explain Ivan Samarskyi author Ondřej Bojar author 2026-10 text Proceedings of the 9th BlackboxNLP Workshop: Analyzing and Interpreting Neural Networks for NLP Dana Arad editor Yonatan Belinkov editor Hanjie Chen editor Nils Feldhus editor Aaron Mueller editor Hosein Mohebbi editor Gabriele Sarti editor Martin Tutek editor Association for Computational Linguistics Budapest, Hungary conference publication 979-8-89176-476-7 samarskyi-bojar-2026-reverse https://aclanthology.org/2026.blackboxnlp-1.23/ 2026-10 374 383 Can Interpretation Predict Behavior on Unseen Data? Victoria R Li author Jenny Kaufmann author Tian Qin author Martin Wattenberg author David Alvarez-Melis author Naomi Saphra author 2026-10 text Proceedings of the 9th BlackboxNLP Workshop: Analyzing and Interpreting Neural Networks for NLP Dana Arad editor Yonatan Belinkov editor Hanjie Chen editor Nils Feldhus editor Aaron Mueller editor Hosein Mohebbi editor Gabriele Sarti editor Martin Tutek editor Association for Computational Linguistics Budapest, Hungary conference publication 979-8-89176-476-7 li-etal-2026-interpretation https://aclanthology.org/2026.blackboxnlp-1.24/ 2026-10 384 407 Multiple Recovery Paths in a Lesioned Transformer: Internal Rerouting and Readout Remapping Lorenzo Cannata author Uri Hasson author 2026-10 text Proceedings of the 9th BlackboxNLP Workshop: Analyzing and Interpreting Neural Networks for NLP Dana Arad editor Yonatan Belinkov editor Hanjie Chen editor Nils Feldhus editor Aaron Mueller editor Hosein Mohebbi editor Gabriele Sarti editor Martin Tutek editor Association for Computational Linguistics Budapest, Hungary conference publication 979-8-89176-476-7 cannata-hasson-2026-multiple https://aclanthology.org/2026.blackboxnlp-1.25/ 2026-10 408 418 Testing the Limits of Truth Directions in LLMs Angelos Poulis author Mark Crovella author Evimaria Terzi author 2026-10 text Proceedings of the 9th BlackboxNLP Workshop: Analyzing and Interpreting Neural Networks for NLP Dana Arad editor Yonatan Belinkov editor Hanjie Chen editor Nils Feldhus editor Aaron Mueller editor Hosein Mohebbi editor Gabriele Sarti editor Martin Tutek editor Association for Computational Linguistics Budapest, Hungary conference publication 979-8-89176-476-7 poulis-etal-2026-testing https://aclanthology.org/2026.blackboxnlp-1.26/ 2026-10 419 447 Forcing a Letter Suppresses Refusal: Mechanistic Evidence from WMDP Anna Antipova author Stepan Epifantsev author Oleg Chernikov author Leon Strelkov author Alexey Dontsov author 2026-10 text Proceedings of the 9th BlackboxNLP Workshop: Analyzing and Interpreting Neural Networks for NLP Dana Arad editor Yonatan Belinkov editor Hanjie Chen editor Nils Feldhus editor Aaron Mueller editor Hosein Mohebbi editor Gabriele Sarti editor Martin Tutek editor Association for Computational Linguistics Budapest, Hungary conference publication 979-8-89176-476-7 antipova-etal-2026-forcing https://aclanthology.org/2026.blackboxnlp-1.27/ 2026-10 448 464 Rewired or Gated? How Instruction Tuning Shapes Knowledge-Conflict Circuits in LLMs Shubham Santosh Pandere author Gautam Ranka author Ritika Varshney author Navya Deshmukh author Roushni Sareen author Roshan Kumar Singh author 2026-10 text Proceedings of the 9th BlackboxNLP Workshop: Analyzing and Interpreting Neural Networks for NLP Dana Arad editor Yonatan Belinkov editor Hanjie Chen editor Nils Feldhus editor Aaron Mueller editor Hosein Mohebbi editor Gabriele Sarti editor Martin Tutek editor Association for Computational Linguistics Budapest, Hungary conference publication 979-8-89176-476-7 pandere-etal-2026-rewired https://aclanthology.org/2026.blackboxnlp-1.28/ 2026-10 465 476 Belief or Circuitry? Causal Evidence for In-Context Graph Learning Katharine Kowalyshyn author Timothy Duggan author Daniel Little author Michael C Hughes author 2026-10 text Proceedings of the 9th BlackboxNLP Workshop: Analyzing and Interpreting Neural Networks for NLP Dana Arad editor Yonatan Belinkov editor Hanjie Chen editor Nils Feldhus editor Aaron Mueller editor Hosein Mohebbi editor Gabriele Sarti editor Martin Tutek editor Association for Computational Linguistics Budapest, Hungary conference publication 979-8-89176-476-7 kowalyshyn-etal-2026-belief https://aclanthology.org/2026.blackboxnlp-1.29/ 2026-10 477 486 Does Out-of-Sight Equal Out-of-Mind in CoT Monitorability? Pedro Lobato Ferreira author Wilker Aziz author Ivan Titov author 2026-10 text Proceedings of the 9th BlackboxNLP Workshop: Analyzing and Interpreting Neural Networks for NLP Dana Arad editor Yonatan Belinkov editor Hanjie Chen editor Nils Feldhus editor Aaron Mueller editor Hosein Mohebbi editor Gabriele Sarti editor Martin Tutek editor Association for Computational Linguistics Budapest, Hungary conference publication 979-8-89176-476-7 ferreira-etal-2026-sight https://aclanthology.org/2026.blackboxnlp-1.30/ 2026-10 487 513 Linear Layer Repair: Residual-Controlled Layer Profiling for Fault-Tolerant Transformer Inference Jiya Sheetal Rathi author André Bauer author 2026-10 text Proceedings of the 9th BlackboxNLP Workshop: Analyzing and Interpreting Neural Networks for NLP Dana Arad editor Yonatan Belinkov editor Hanjie Chen editor Nils Feldhus editor Aaron Mueller editor Hosein Mohebbi editor Gabriele Sarti editor Martin Tutek editor Association for Computational Linguistics Budapest, Hungary conference publication 979-8-89176-476-7 rathi-bauer-2026-linear https://aclanthology.org/2026.blackboxnlp-1.31/ 2026-10 514 522 Pitfalls in Evaluating Interpretability Agents Tal Haklay author Nikhil Prakash author Sana Pandey author Antonio Torralba author Aaron Mueller author Jacob Andreas author Tamar Rott Shaham author Yonatan Belinkov author 2026-10 text Proceedings of the 9th BlackboxNLP Workshop: Analyzing and Interpreting Neural Networks for NLP Dana Arad editor Yonatan Belinkov editor Hanjie Chen editor Nils Feldhus editor Aaron Mueller editor Hosein Mohebbi editor Gabriele Sarti editor Martin Tutek editor Association for Computational Linguistics Budapest, Hungary conference publication 979-8-89176-476-7 haklay-etal-2026-pitfalls https://aclanthology.org/2026.blackboxnlp-1.32/ 2026-10 523 560 Zipf Made Me Do It: A Causal Account of Outlier Dimensions Kartik Sharma author Shreshth Rai author Nikunj Gupta author Dinesh K Vishwakarma author 2026-10 text Proceedings of the 9th BlackboxNLP Workshop: Analyzing and Interpreting Neural Networks for NLP Dana Arad editor Yonatan Belinkov editor Hanjie Chen editor Nils Feldhus editor Aaron Mueller editor Hosein Mohebbi editor Gabriele Sarti editor Martin Tutek editor Association for Computational Linguistics Budapest, Hungary conference publication 979-8-89176-476-7 sharma-etal-2026-zipf https://aclanthology.org/2026.blackboxnlp-1.33/ 2026-10 561 572 Are Arithmetic Heuristic Neurons Form-Invariant? A Mechanistic Analysis of Symbols, Text, and Code in LLMs Sharath Naganna author Tanvir Ahmed Sijan author Uddipta Kalita author 2026-10 text Proceedings of the 9th BlackboxNLP Workshop: Analyzing and Interpreting Neural Networks for NLP Dana Arad editor Yonatan Belinkov editor Hanjie Chen editor Nils Feldhus editor Aaron Mueller editor Hosein Mohebbi editor Gabriele Sarti editor Martin Tutek editor Association for Computational Linguistics Budapest, Hungary conference publication 979-8-89176-476-7 naganna-etal-2026-arithmetic https://aclanthology.org/2026.blackboxnlp-1.34/ 2026-10 573 589 A Multi-Dimensional Evaluation of Explainability in Media Bias Detection Ting Chen author Chengjun Zhang author Benjamin M Ampel author Sagar Samtani author 2026-10 text Proceedings of the 9th BlackboxNLP Workshop: Analyzing and Interpreting Neural Networks for NLP Dana Arad editor Yonatan Belinkov editor Hanjie Chen editor Nils Feldhus editor Aaron Mueller editor Hosein Mohebbi editor Gabriele Sarti editor Martin Tutek editor Association for Computational Linguistics Budapest, Hungary conference publication 979-8-89176-476-7 chen-etal-2026-multi-dimensional https://aclanthology.org/2026.blackboxnlp-1.35/ 2026-10 590 602 Automatically Interpreting Attribution Graphs via Probe Prompting Giuseppe Birardi author Gonçalo Paulo author 2026-10 text Proceedings of the 9th BlackboxNLP Workshop: Analyzing and Interpreting Neural Networks for NLP Dana Arad editor Yonatan Belinkov editor Hanjie Chen editor Nils Feldhus editor Aaron Mueller editor Hosein Mohebbi editor Gabriele Sarti editor Martin Tutek editor Association for Computational Linguistics Budapest, Hungary conference publication 979-8-89176-476-7 birardi-paulo-2026-automatically https://aclanthology.org/2026.blackboxnlp-1.36/ 2026-10 603 638 Sampling Reveals Style: Unsupervised, Training-Free Discovery of Prompt-Conditional Stylistic Axes in LLM Activations Ajit Mallavarapu author Ziwei Gu author 2026-10 text Proceedings of the 9th BlackboxNLP Workshop: Analyzing and Interpreting Neural Networks for NLP Dana Arad editor Yonatan Belinkov editor Hanjie Chen editor Nils Feldhus editor Aaron Mueller editor Hosein Mohebbi editor Gabriele Sarti editor Martin Tutek editor Association for Computational Linguistics Budapest, Hungary conference publication 979-8-89176-476-7 mallavarapu-gu-2026-sampling https://aclanthology.org/2026.blackboxnlp-1.37/ 2026-10 639 654 Compositional Multilingual and Behavioral Attribute Steering Hyun Gu Kang author Daniil Gurgurov author Tanja Baeumel author Josef Van Genabith author Simon Ostermann author 2026-10 text Proceedings of the 9th BlackboxNLP Workshop: Analyzing and Interpreting Neural Networks for NLP Dana Arad editor Yonatan Belinkov editor Hanjie Chen editor Nils Feldhus editor Aaron Mueller editor Hosein Mohebbi editor Gabriele Sarti editor Martin Tutek editor Association for Computational Linguistics Budapest, Hungary conference publication 979-8-89176-476-7 kang-etal-2026-compositional https://aclanthology.org/2026.blackboxnlp-1.38/ 2026-10 655 676 Are XAI Metrics Meaningful Proxies for Human Judgment? Evaluation Explanation Quality in Claim Matching Jaroslav Kopcan author Ľuboš Kriš author Qiwei Peng author Miroslav Blstak author Vera Schmitt author 2026-10 text Proceedings of the 9th BlackboxNLP Workshop: Analyzing and Interpreting Neural Networks for NLP Dana Arad editor Yonatan Belinkov editor Hanjie Chen editor Nils Feldhus editor Aaron Mueller editor Hosein Mohebbi editor Gabriele Sarti editor Martin Tutek editor Association for Computational Linguistics Budapest, Hungary conference publication 979-8-89176-476-7 kopcan-etal-2026-xai https://aclanthology.org/2026.blackboxnlp-1.39/ 2026-10 677 695 Your Judge Inherits Your Prompt: In-Context Label Bias in an LLM-as-Judge Is Directional, Not Symmetric Abdullah Ahmad author Ahmed J O Ahmed author Luca Citi author 2026-10 text Proceedings of the 9th BlackboxNLP Workshop: Analyzing and Interpreting Neural Networks for NLP Dana Arad editor Yonatan Belinkov editor Hanjie Chen editor Nils Feldhus editor Aaron Mueller editor Hosein Mohebbi editor Gabriele Sarti editor Martin Tutek editor Association for Computational Linguistics Budapest, Hungary conference publication 979-8-89176-476-7 ahmad-etal-2026-judge https://aclanthology.org/2026.blackboxnlp-1.40/ 2026-10 696 706 Attend or Perish: Benchmarking Attention on Algorithmic Reasoning Michal Spiegel author Michal Štefánik author Marek Kadlčík author Valeria Ruscio author Martin Fajčík author Josef Kuchař author 2026-10 text Proceedings of the 9th BlackboxNLP Workshop: Analyzing and Interpreting Neural Networks for NLP Dana Arad editor Yonatan Belinkov editor Hanjie Chen editor Nils Feldhus editor Aaron Mueller editor Hosein Mohebbi editor Gabriele Sarti editor Martin Tutek editor Association for Computational Linguistics Budapest, Hungary conference publication 979-8-89176-476-7 spiegel-etal-2026-attend https://aclanthology.org/2026.blackboxnlp-1.41/ 2026-10 707 725 Dissecting Lie Expectancy in Language Models Xingyuan Zhao author Seana Coulson author 2026-10 text Proceedings of the 9th BlackboxNLP Workshop: Analyzing and Interpreting Neural Networks for NLP Dana Arad editor Yonatan Belinkov editor Hanjie Chen editor Nils Feldhus editor Aaron Mueller editor Hosein Mohebbi editor Gabriele Sarti editor Martin Tutek editor Association for Computational Linguistics Budapest, Hungary conference publication 979-8-89176-476-7 zhao-coulson-2026-dissecting https://aclanthology.org/2026.blackboxnlp-1.42/ 2026-10 726 738 The Geometry of Political Abstraction in Large Language Models Noé Durandard author Saurabh Dhawan author Thierry Poibeau author 2026-10 text Proceedings of the 9th BlackboxNLP Workshop: Analyzing and Interpreting Neural Networks for NLP Dana Arad editor Yonatan Belinkov editor Hanjie Chen editor Nils Feldhus editor Aaron Mueller editor Hosein Mohebbi editor Gabriele Sarti editor Martin Tutek editor Association for Computational Linguistics Budapest, Hungary conference publication 979-8-89176-476-7 durandard-etal-2026-geometry https://aclanthology.org/2026.blackboxnlp-1.43/ 2026-10 739 760 How Much Sonority Structure Do Phonotactic Language Models Learn from Distribution? Jia Can Xu author Mario Giulianelli author 2026-10 text Proceedings of the 9th BlackboxNLP Workshop: Analyzing and Interpreting Neural Networks for NLP Dana Arad editor Yonatan Belinkov editor Hanjie Chen editor Nils Feldhus editor Aaron Mueller editor Hosein Mohebbi editor Gabriele Sarti editor Martin Tutek editor Association for Computational Linguistics Budapest, Hungary conference publication 979-8-89176-476-7 xu-giulianelli-2026-much https://aclanthology.org/2026.blackboxnlp-1.44/ 2026-10 761 782 Do ICL and Finetuned Misalignment Share a Causal Component? Evidence from Risky Financial Advice Albrave Raphael Albayda author 2026-10 text Proceedings of the 9th BlackboxNLP Workshop: Analyzing and Interpreting Neural Networks for NLP Dana Arad editor Yonatan Belinkov editor Hanjie Chen editor Nils Feldhus editor Aaron Mueller editor Hosein Mohebbi editor Gabriele Sarti editor Martin Tutek editor Association for Computational Linguistics Budapest, Hungary conference publication 979-8-89176-476-7 albayda-2026-icl https://aclanthology.org/2026.blackboxnlp-1.45/ 2026-10 783 794 From Rage to Refusal : Using Social-Media Data to Steer Hostility-Related Affect in LLMs Gayatri Krishnakumar author Maximilian Puelma Touzel author Zachary Yang author Jean-François Godbout author Reihaneh Rabbany author 2026-10 text Proceedings of the 9th BlackboxNLP Workshop: Analyzing and Interpreting Neural Networks for NLP Dana Arad editor Yonatan Belinkov editor Hanjie Chen editor Nils Feldhus editor Aaron Mueller editor Hosein Mohebbi editor Gabriele Sarti editor Martin Tutek editor Association for Computational Linguistics Budapest, Hungary conference publication 979-8-89176-476-7 krishnakumar-etal-2026-rage https://aclanthology.org/2026.blackboxnlp-1.46/ 2026-10 795 817 Where Does Authorship Signal Emerge in Encoder-Based Language Models? Francis Kulumba author Guillaume Vimont author Laurent Romary author Florian Cafiero author 2026-10 text Proceedings of the 9th BlackboxNLP Workshop: Analyzing and Interpreting Neural Networks for NLP Dana Arad editor Yonatan Belinkov editor Hanjie Chen editor Nils Feldhus editor Aaron Mueller editor Hosein Mohebbi editor Gabriele Sarti editor Martin Tutek editor Association for Computational Linguistics Budapest, Hungary conference publication 979-8-89176-476-7 kulumba-etal-2026-authorship https://aclanthology.org/2026.blackboxnlp-1.47/ 2026-10 818 834 A Survey on the Linear Representation Hypothesis Sewoong Lee author Marc E Canby author Ikhyun Cho author Julia Hockenmaier author 2026-10 text Proceedings of the 9th BlackboxNLP Workshop: Analyzing and Interpreting Neural Networks for NLP Dana Arad editor Yonatan Belinkov editor Hanjie Chen editor Nils Feldhus editor Aaron Mueller editor Hosein Mohebbi editor Gabriele Sarti editor Martin Tutek editor Association for Computational Linguistics Budapest, Hungary conference publication 979-8-89176-476-7 lee-etal-2026-survey https://aclanthology.org/2026.blackboxnlp-1.48/ 2026-10 835 851 The Latin Substrate: How Language Models Represent and Mediate Script Choice Daniil Gurgurov author Alan Saji author Katharina Trinley author Josef Van Genabith author Simon Ostermann author 2026-10 text Proceedings of the 9th BlackboxNLP Workshop: Analyzing and Interpreting Neural Networks for NLP Dana Arad editor Yonatan Belinkov editor Hanjie Chen editor Nils Feldhus editor Aaron Mueller editor Hosein Mohebbi editor Gabriele Sarti editor Martin Tutek editor Association for Computational Linguistics Budapest, Hungary conference publication 979-8-89176-476-7 gurgurov-etal-2026-latin https://aclanthology.org/2026.blackboxnlp-1.49/ 2026-10 852 878 MACE-XAI: A Multi-Agent Crowd Evaluation Framework for Robust Natural Language Explanations Renato Okabayashi Miyaji author 2026-10 text Proceedings of the 9th BlackboxNLP Workshop: Analyzing and Interpreting Neural Networks for NLP Dana Arad editor Yonatan Belinkov editor Hanjie Chen editor Nils Feldhus editor Aaron Mueller editor Hosein Mohebbi editor Gabriele Sarti editor Martin Tutek editor Association for Computational Linguistics Budapest, Hungary conference publication 979-8-89176-476-7 miyaji-2026-mace https://aclanthology.org/2026.blackboxnlp-1.50/ 2026-10 879 891 Automated Taxonomy Discovery for Interpretable Diagnosis of Multi-Hop Question Answering Datasets Ryoji Osaki author Masaki Asada author Makoto Miwa author 2026-10 text Proceedings of the 9th BlackboxNLP Workshop: Analyzing and Interpreting Neural Networks for NLP Dana Arad editor Yonatan Belinkov editor Hanjie Chen editor Nils Feldhus editor Aaron Mueller editor Hosein Mohebbi editor Gabriele Sarti editor Martin Tutek editor Association for Computational Linguistics Budapest, Hungary conference publication 979-8-89176-476-7 osaki-etal-2026-automated https://aclanthology.org/2026.blackboxnlp-1.51/ 2026-10 892 918 Explaining the Influence of Nonverbal Streams Across Sentence Types Lipisha Chaudhary author Ifeoma Nwogu author 2026-10 text Proceedings of the 9th BlackboxNLP Workshop: Analyzing and Interpreting Neural Networks for NLP Dana Arad editor Yonatan Belinkov editor Hanjie Chen editor Nils Feldhus editor Aaron Mueller editor Hosein Mohebbi editor Gabriele Sarti editor Martin Tutek editor Association for Computational Linguistics Budapest, Hungary conference publication 979-8-89176-476-7 chaudhary-nwogu-2026-explaining https://aclanthology.org/2026.blackboxnlp-1.52/ 2026-10 919 936 MERGE: Minimal Expression-Replacement GEneralization Test for Natural Language Inference Mădălina Zgreabăn author Tejaswini Deoskar author Lasha Abzianidze author 2026-10 text Proceedings of the 9th BlackboxNLP Workshop: Analyzing and Interpreting Neural Networks for NLP Dana Arad editor Yonatan Belinkov editor Hanjie Chen editor Nils Feldhus editor Aaron Mueller editor Hosein Mohebbi editor Gabriele Sarti editor Martin Tutek editor Association for Computational Linguistics Budapest, Hungary conference publication 979-8-89176-476-7 zgreaban-etal-2026-merge https://aclanthology.org/2026.blackboxnlp-1.53/ 2026-10 937 964 The Geometry of Metaphor: Probing and Steering Metaphor Representations in Large Language Models Shihui Li author Mette Friis Andersen author Martha Lewis author 2026-10 text Proceedings of the 9th BlackboxNLP Workshop: Analyzing and Interpreting Neural Networks for NLP Dana Arad editor Yonatan Belinkov editor Hanjie Chen editor Nils Feldhus editor Aaron Mueller editor Hosein Mohebbi editor Gabriele Sarti editor Martin Tutek editor Association for Computational Linguistics Budapest, Hungary conference publication 979-8-89176-476-7 li-etal-2026-geometry https://aclanthology.org/2026.blackboxnlp-1.54/ 2026-10 965 985 When Context Misleads: Surprisal, Energy and Attention Entropy as Metrics of Coherence Illusions in LLMs Ece Takmaz author Nitin Kumar author Li Kloostra author Jakub Dotlacil author 2026-10 text Proceedings of the 9th BlackboxNLP Workshop: Analyzing and Interpreting Neural Networks for NLP Dana Arad editor Yonatan Belinkov editor Hanjie Chen editor Nils Feldhus editor Aaron Mueller editor Hosein Mohebbi editor Gabriele Sarti editor Martin Tutek editor Association for Computational Linguistics Budapest, Hungary conference publication 979-8-89176-476-7 takmaz-etal-2026-context https://aclanthology.org/2026.blackboxnlp-1.55/ 2026-10 986 1006 Present but Barely Used: Reproducing the Emergent Othello World Model in a Masked Diffusion Language Model Éric Jacopin author 2026-10 text Proceedings of the 9th BlackboxNLP Workshop: Analyzing and Interpreting Neural Networks for NLP Dana Arad editor Yonatan Belinkov editor Hanjie Chen editor Nils Feldhus editor Aaron Mueller editor Hosein Mohebbi editor Gabriele Sarti editor Martin Tutek editor Association for Computational Linguistics Budapest, Hungary conference publication 979-8-89176-476-7 jacopin-2026-present https://aclanthology.org/2026.blackboxnlp-1.56/ 2026-10 1007 1016 Do Function Vectors and Task Vectors Generalize? A Reproducibility Study Across Model Families and Scale Serendeep Rudraraju author 2026-10 text Proceedings of the 9th BlackboxNLP Workshop: Analyzing and Interpreting Neural Networks for NLP Dana Arad editor Yonatan Belinkov editor Hanjie Chen editor Nils Feldhus editor Aaron Mueller editor Hosein Mohebbi editor Gabriele Sarti editor Martin Tutek editor Association for Computational Linguistics Budapest, Hungary conference publication 979-8-89176-476-7 rudraraju-2026-function https://aclanthology.org/2026.blackboxnlp-1.57/ 2026-10 1017 1028 Memory vs. Context? Influential Factors of Factual Recall in Language Models Guilhem Fouilhé author Nicholas Asher author Philippe Muller author 2026-10 text Proceedings of the 9th BlackboxNLP Workshop: Analyzing and Interpreting Neural Networks for NLP Dana Arad editor Yonatan Belinkov editor Hanjie Chen editor Nils Feldhus editor Aaron Mueller editor Hosein Mohebbi editor Gabriele Sarti editor Martin Tutek editor Association for Computational Linguistics Budapest, Hungary conference publication 979-8-89176-476-7 fouilhe-etal-2026-memory https://aclanthology.org/2026.blackboxnlp-1.58/ 2026-10 1029 1046 Demystifying Entropy-based Selection for Chain-of-Thought Compression in Large Reasoning Models Sara Candussio author Daniel Scalena author Elisabetta Fersini author Luca Bortolussi author Malvina Nissim author Gabriele Sarti author 2026-10 text Proceedings of the 9th BlackboxNLP Workshop: Analyzing and Interpreting Neural Networks for NLP Dana Arad editor Yonatan Belinkov editor Hanjie Chen editor Nils Feldhus editor Aaron Mueller editor Hosein Mohebbi editor Gabriele Sarti editor Martin Tutek editor Association for Computational Linguistics Budapest, Hungary conference publication 979-8-89176-476-7 candussio-etal-2026-demystifying https://aclanthology.org/2026.blackboxnlp-1.59/ 2026-10 1047 1066 Limitations of Automated Simulatability: LLM Simulators Can Bypass Explanations Antonin Poché author Fanny Jourdan author Nils Feldhus author Qianli Wang author Jing Yang author Simon Ostermann author Nicholas Asher author Philippe Muller author Vera Schmitt author 2026-10 text Proceedings of the 9th BlackboxNLP Workshop: Analyzing and Interpreting Neural Networks for NLP Dana Arad editor Yonatan Belinkov editor Hanjie Chen editor Nils Feldhus editor Aaron Mueller editor Hosein Mohebbi editor Gabriele Sarti editor Martin Tutek editor Association for Computational Linguistics Budapest, Hungary conference publication 979-8-89176-476-7 poche-etal-2026-limitations https://aclanthology.org/2026.blackboxnlp-1.60/ 2026-10 1067 1099 Recurrence Is Not Enough: Causally Validating Multilingual SAE Translation Features in Gemma 2 and 3 Giang Son Nguyen author Nhi Ngoc-Yen Nguyen author Wray Buntine author Dung D Le author 2026-10 text Proceedings of the 9th BlackboxNLP Workshop: Analyzing and Interpreting Neural Networks for NLP Dana Arad editor Yonatan Belinkov editor Hanjie Chen editor Nils Feldhus editor Aaron Mueller editor Hosein Mohebbi editor Gabriele Sarti editor Martin Tutek editor Association for Computational Linguistics Budapest, Hungary conference publication 979-8-89176-476-7 nguyen-etal-2026-recurrence https://aclanthology.org/2026.blackboxnlp-1.61/ 2026-10 1100 1112 Reproducing and Evaluating the Generalizability of Subliminal Learning in Open-Weight Models Daan van der Weijden author Nathan Brack author Selene Baez Santamaria author 2026-10 text Proceedings of the 9th BlackboxNLP Workshop: Analyzing and Interpreting Neural Networks for NLP Dana Arad editor Yonatan Belinkov editor Hanjie Chen editor Nils Feldhus editor Aaron Mueller editor Hosein Mohebbi editor Gabriele Sarti editor Martin Tutek editor Association for Computational Linguistics Budapest, Hungary conference publication 979-8-89176-476-7 van-der-weijden-etal-2026-reproducing https://aclanthology.org/2026.blackboxnlp-1.62/ 2026-10 1113 1120 Can Sparse AutoEncoder-Based Interpretability Demonstrate Geographic Knowledge Beyond Orthographic Features? Thomas Hitchon author Rémy Decoupes author Adrien Guille author 2026-10 text Proceedings of the 9th BlackboxNLP Workshop: Analyzing and Interpreting Neural Networks for NLP Dana Arad editor Yonatan Belinkov editor Hanjie Chen editor Nils Feldhus editor Aaron Mueller editor Hosein Mohebbi editor Gabriele Sarti editor Martin Tutek editor Association for Computational Linguistics Budapest, Hungary conference publication 979-8-89176-476-7 hitchon-etal-2026-sparse https://aclanthology.org/2026.blackboxnlp-1.63/ 2026-10 1121 1127 Inoculate or Reflect? Two Training Interventions Under Prompting, Steering, and Patching Ayesha Imran author Muhammad Aaliyan author 2026-10 text Proceedings of the 9th BlackboxNLP Workshop: Analyzing and Interpreting Neural Networks for NLP Dana Arad editor Yonatan Belinkov editor Hanjie Chen editor Nils Feldhus editor Aaron Mueller editor Hosein Mohebbi editor Gabriele Sarti editor Martin Tutek editor Association for Computational Linguistics Budapest, Hungary conference publication 979-8-89176-476-7 imran-aaliyan-2026-inoculate https://aclanthology.org/2026.blackboxnlp-1.64/ 2026-10 1128 1142 Is Refusal Direction Universal across Low-Resourced Languages? Anastasia Ivanova author 2026-10 text Proceedings of the 9th BlackboxNLP Workshop: Analyzing and Interpreting Neural Networks for NLP Dana Arad editor Yonatan Belinkov editor Hanjie Chen editor Nils Feldhus editor Aaron Mueller editor Hosein Mohebbi editor Gabriele Sarti editor Martin Tutek editor Association for Computational Linguistics Budapest, Hungary conference publication 979-8-89176-476-7 ivanova-2026-refusal https://aclanthology.org/2026.blackboxnlp-1.65/ 2026-10 1143 1154 Scale-Invariant Normalization Hides Activation Scaling: Why Patch Site Matters for Demographic Bias Localization in Clinical Text Generation Yoana Tsoneva author Fedor Splitt author Awsam Agbaria author Alexandra Tsiakalou author Jing Yang author Vera Schmitt author 2026-10 text Proceedings of the 9th BlackboxNLP Workshop: Analyzing and Interpreting Neural Networks for NLP Dana Arad editor Yonatan Belinkov editor Hanjie Chen editor Nils Feldhus editor Aaron Mueller editor Hosein Mohebbi editor Gabriele Sarti editor Martin Tutek editor Association for Computational Linguistics Budapest, Hungary conference publication 979-8-89176-476-7 tsoneva-etal-2026-scale https://aclanthology.org/2026.blackboxnlp-1.66/ 2026-10 1155 1166 Probing Without Labels: Contrast-only Reproductions of Interpretability Results Stefan F Schouten author Peter Bloem author 2026-10 text Proceedings of the 9th BlackboxNLP Workshop: Analyzing and Interpreting Neural Networks for NLP Dana Arad editor Yonatan Belinkov editor Hanjie Chen editor Nils Feldhus editor Aaron Mueller editor Hosein Mohebbi editor Gabriele Sarti editor Martin Tutek editor Association for Computational Linguistics Budapest, Hungary conference publication 979-8-89176-476-7 schouten-bloem-2026-probing https://aclanthology.org/2026.blackboxnlp-1.67/ 2026-10 1167 1183 Reproducing the Results, But not the Interpretations: Evidence for Artifacts in Automatic Creativity Scoring with LLMs Carl Hermany author Omar Momen author Sina Zarrieß author 2026-10 text Proceedings of the 9th BlackboxNLP Workshop: Analyzing and Interpreting Neural Networks for NLP Dana Arad editor Yonatan Belinkov editor Hanjie Chen editor Nils Feldhus editor Aaron Mueller editor Hosein Mohebbi editor Gabriele Sarti editor Martin Tutek editor Association for Computational Linguistics Budapest, Hungary conference publication 979-8-89176-476-7 hermany-etal-2026-reproducing https://aclanthology.org/2026.blackboxnlp-1.68/ 2026-10 1184 1195 Claude reinvestigates herself: A replication of different types of syntactic agreement recruit the same units in LLMs Elena Polyakova author Shivan Arora author Suchir Salhan author Paula Buttery author 2026-10 text Proceedings of the 9th BlackboxNLP Workshop: Analyzing and Interpreting Neural Networks for NLP Dana Arad editor Yonatan Belinkov editor Hanjie Chen editor Nils Feldhus editor Aaron Mueller editor Hosein Mohebbi editor Gabriele Sarti editor Martin Tutek editor Association for Computational Linguistics Budapest, Hungary conference publication 979-8-89176-476-7 polyakova-etal-2026-claude https://aclanthology.org/2026.blackboxnlp-1.69/ 2026-10 1196 1222 Are Latent Reasoning Tokens Unnecessary for Model Performance? Austin Long author Colin Paterson author Jing Xu author 2026-10 text Proceedings of the 9th BlackboxNLP Workshop: Analyzing and Interpreting Neural Networks for NLP Dana Arad editor Yonatan Belinkov editor Hanjie Chen editor Nils Feldhus editor Aaron Mueller editor Hosein Mohebbi editor Gabriele Sarti editor Martin Tutek editor Association for Computational Linguistics Budapest, Hungary conference publication 979-8-89176-476-7 long-etal-2026-latent https://aclanthology.org/2026.blackboxnlp-1.70/ 2026-10 1223 1232 Same Sink, Different Plumbing: Reproducing and Refining the Mechanism of Attention Sinks Abdullah Al Zubayer author Syed Naveed Mahmood author Md Rezaur Rahman Bhuiyan author Mir Md Mohiuddin Abrar author Jareen Tasneem Khondaker author Tasfia Zaman author Farig Sadeque author 2026-10 text Proceedings of the 9th BlackboxNLP Workshop: Analyzing and Interpreting Neural Networks for NLP Dana Arad editor Yonatan Belinkov editor Hanjie Chen editor Nils Feldhus editor Aaron Mueller editor Hosein Mohebbi editor Gabriele Sarti editor Martin Tutek editor Association for Computational Linguistics Budapest, Hungary conference publication 979-8-89176-476-7 al-zubayer-etal-2026-sink https://aclanthology.org/2026.blackboxnlp-1.71/ 2026-10 1233 1247 Development of Linear Truth Encodings in Language Models: A Replication Study Mingchuan Zou author Suchir Salhan author Paula Buttery author 2026-10 text Proceedings of the 9th BlackboxNLP Workshop: Analyzing and Interpreting Neural Networks for NLP Dana Arad editor Yonatan Belinkov editor Hanjie Chen editor Nils Feldhus editor Aaron Mueller editor Hosein Mohebbi editor Gabriele Sarti editor Martin Tutek editor Association for Computational Linguistics Budapest, Hungary conference publication 979-8-89176-476-7 zou-etal-2026-development https://aclanthology.org/2026.blackboxnlp-1.72/ 2026-10 1248 1263 Stress-Testing Random-Vector Controls for Activation Steering Behnood Mashayekhi author Parsa Hosseini author Arian Amiramjadi author 2026-10 text Proceedings of the 9th BlackboxNLP Workshop: Analyzing and Interpreting Neural Networks for NLP Dana Arad editor Yonatan Belinkov editor Hanjie Chen editor Nils Feldhus editor Aaron Mueller editor Hosein Mohebbi editor Gabriele Sarti editor Martin Tutek editor Association for Computational Linguistics Budapest, Hungary conference publication 979-8-89176-476-7 mashayekhi-etal-2026-stress https://aclanthology.org/2026.blackboxnlp-1.73/ 2026-10 1264 1272 Reproducing Transparent and Scrutable Recommendations: Exploring Open-Weight Models via Natural-Language User Profiles Noah Mamié author Laurin van den Bergh author 2026-10 text Proceedings of the 9th BlackboxNLP Workshop: Analyzing and Interpreting Neural Networks for NLP Dana Arad editor Yonatan Belinkov editor Hanjie Chen editor Nils Feldhus editor Aaron Mueller editor Hosein Mohebbi editor Gabriele Sarti editor Martin Tutek editor Association for Computational Linguistics Budapest, Hungary conference publication 979-8-89176-476-7 mamie-van-den-bergh-2026-reproducing https://aclanthology.org/2026.blackboxnlp-1.74/ 2026-10 1273 1284