Accepted Papers BlackboxNLP 2024

Full Papers

• Accelerating Sparse Autoencoder Training via Layer-Wise Transfer Learning in Large Language Models
  Davide Ghilardi, Federico Belotti, Marco Molinari, Jaehyuk Lim

• Toward the Evaluation of Large Language Models Considering Score Variance across Instruction Templates
  Yusuke Sakai, Adam Nohejl, JIANGNAN HANG, Hidetaka Kamigaito, Taro Watanabe

• Mechanistic?
  Sarah Wiegreffe, Naomi Saphra

• Investigating Layer Importance in Large Language Models
  Yang Zhang, Yanfei Dong, Kenji Kawaguchi

• Counterfactuals As a Means for Evaluating Faithfulness of Attribution Methods in Autoregressive Language Models
  Sepehr Kamahi, Yadollah Yaghoobzadeh

• IvRA: A Framework to Enhance Attention-Based Explanations for Language Models with Interpretability-Driven Training
  Sean Xie, Soroush Vosoughi, Saeed Hassanpour

• Pruning for Protection: Increasing Jailbreak Resistance in Aligned LLMs Without Fine-Tuning
  Adib Hasan, Ileana Rugina, Alex Wang

• Attribution Patching Outperforms Automated Circuit Discovery
  Aaquib Syed, Can Rager, Arthur Conmy

• Do Metadata and Appearance of the Retrieved Webpages Affect LLM’s Reasoning in Retrieval-Augmented Generation?
  Cheng-Han Chiang, Hung-yi Lee

• Interpretable by Design: Wrapper Boxes Combine Neural Performance with Faithful Attribution of Model Decisions to Training Data
  Yiheng Su, Junyi Jessy Li, Matthew Lease

• WellDunn: On the Robustness and Explainability of Language Models and Large Language Models in Identifying Wellness Dimensions
  Seyedali Mohammadi, Edward Raff, Jinendra Malekar, Vedant Palit, Francis Ferraro, Manas Gaur

• Copy Suppression: Comprehensively Understanding a Motif in Language Model Attention Heads
  Callum Stuart McDougall, Arthur Conmy, Cody Rushing, Thomas McGrath, Neel Nanda

• How Language Models Prioritize Contextual Grammatical Cues?
  Hamidreza Amirzadeh, Afra Alishahi, Hosein Mohebbi

• Self-Assessment Tests are Unreliable Measures of LLM Personality
  Akshat Gupta, Xiaoyang Song, Gopala Anumanchipalli

• Gemma Scope: Open Sparse Autoencoders Everywhere All At Once on Gemma 2
  Tom Lieberum, Senthooran Rajamanoharan, Arthur Conmy, Lewis Smith, Nicolas Sonnerat, Vikrant Varma, Janos Kramar, Anca Dragan, Rohin Shah, Neel Nanda

• Log Probabilities Are a Reliable Estimate of Semantic Plausibility in Base and Instruction-Tuned Language Models
  Carina Kauf, Emmanuele Chersoni, Alessandro Lenci, Evelina Fedorenko, Anna A Ivanova

• Recurrent Neural Networks Learn to Store and Generate Sequences using Non-Linear Representations
  Róbert Csordás, Christopher Potts, Christopher D Manning, Atticus Geiger

• Uncovering Syllable Constituents in the Self-Attention-Based Speech Representations of Whisper
  Erfan A Shams, Iona Gessinger, Julie Carson-Berndsen

• An Adversarial Example for Direct Logit Attribution: Memory Management in GELU-4L
  Jett Janiak, Can Rager, James Dao, Yeu-Tong Lau

• On the alignment of LM language generation and human language comprehension
  Lena Sophia Bolliger, Patrick Haller, Lena Ann Jäger

• Transformers Learn Hidden Dynamics when Trained to Predict Markov Decision Processes
  Yuxi Chen, Suwei Ma, Tony Dear, Xu Chen

• Faithfulness and the Notion of Adversarial Sensitivity in NLP Explanations
  Supriya Manna, Niladri Sett

• Enhancing Question Answering on Charts Through Effective Pre-training Tasks
  Ashim Gupta, Vivek Gupta, Shuo Zhang, Yujie He, Ning Zhang, Shalin Shah

• Attend First, Consolidate Later: On the Importance of Attention in Different LLM Layers
  Amit Ben Artzy, Roy Schwartz

• Can We Statically Locate Knowledge in Large Language Models? Financial Domain and Toxicity Reduction Case Studies
  Jordi Armengol-Estapé, Lingyu Li, Sebastian Gehrmann, Achintya Gopal, David S Rosenberg, Gideon S. Mann, Mark Dredze

• MultiContrievers: Analysis of Dense Retrieval Representations
  Seraphina Goldfarb-Tarrant, Pedro Rodriguez, Jane Dwivedi-Yu, Patrick Lewis

• Enhancing adversarial robustness in Natural Language Inference using explanations
  Alexandros Koulakos, Maria Lymperaiou, Giorgos Filandrianos, Giorgos Stamou

• LLM Internal States Reveal Hallucination Risk Faced With a Query
  Ziwei Ji, Delong Chen, Etsuko Ishii, Samuel Cahyawijaya, Yejin Bang, Bryan Wilie, Pascale Fung

• Language Models Linearly Represent Sentiment
  Oskar John Hollinsworth, Curt Tigges, Atticus Geiger, Neel Nanda

• Learning, Forgetting, Remembering: Insights From Tracking LLM Memorization During Training
  Danny D. Leybzon, Corentin Kervadec

• Are there identifiable structural parts in the sentence embedding whole?
  Vivi Nastase, Paola Merlo

• Routing in Sparsely-gated Language Models responds to Context
  Stefan Arnold, Marian Fietta, Dilara Yesilbas

• Optimal and efficient text counterfactuals using Graph Neural Networks
  Dimitris Lymperopoulos, Maria Lymperaiou, Giorgos Filandrianos, Giorgos Stamou

• Probing Language Models on Their Knowledge Source
  Zineddine Tighidet, Jiali Mei, Benjamin Piwowarski, Patrick Gallinari

• Multi-property Steering of Large Language Models with Dynamic Activation Composition
  Daniel Scalena, Gabriele Sarti, Malvina Nissim

Non-archival extended abstracts

• Fifty shapes of BLiMP: syntactic learning curves in language models are not uniform, but sometimes unruly
  Bastian Bunzeck, Sina Zarrieß

• Competition of Mechanisms: Tracing How Language Models Handle Facts and Counterfactuals
  Francesco Ortu, Zhijing Jin, Diego Doimo, Mrinmaya Sachan, Alberto Cazzaniga, Bernhard Schölkopf

• Inducing Induction in Llama via Linear Probe Interventions
  Sheridan Feucht, Byron C Wallace, David Bau

• Implicit Meta-Learning in Small Transformer Models: Insights from a Toy Task
  Luan Fletcher, Victor Levoso, Kunvar Thaman, Misha Kilianovski

• Latent Concept-based Explanation of NLP Models
  Xuemin Yu, Fahim Dalvi, Nadir Durrani, Marzia Nouri, Hassan Sajjad

• Exploring Alignment in Shared Cross-Lingual Spaces
  Basel Mousi, Nadir Durrani, Fahim Dalvi, Majd Hawasly, Ahmed Abdelali

• Compositional Cores: Persistent Attention Patterns in Compositionally Generalizing Subnetworks
  Michael Y. Hu, Chuan Shi, Tal Linzen

• How LLMs Reinforce Political Misinformation: Insights from the Analysis of False Presuppositions
  Judith Sieker, Clara Lachenmaier, Sina Zarrieß

• Does Alignment Tuning Really Break LLMs’ Internal Confidence?
  Hongseok Oh, Wonseok Hwang

• How Does Code Pretraining Affect Language Model Task Performance?
  Jackson Petty, Sjoerd van Steenkiste, Tal Linzen

• ToxiSight: Insights Towards Detected Chat Toxicity
  Zachary Yang, Domenico Tullo, Reihaneh Rabbany

• Clusters Emerge in Transformer-based Causal Language Models
  Xinbo Wu, Lav R. Varshney

• Quantifying reliance on external information over parametric knowledge during Retrieval Augmented Generation (RAG) using mechanistic analysis
  Reshmi Ghosh, Rahul Seetharaman, Hitesh Wadhwa, Somyaa Aggarwal, Samyadeep Basu, Soundararajan Srinivasan, Wenlong Zhao, Shreyas Chaudhari, Ehsan Aghazadeh

• Mind Your Manners: Detoxifying Language Models via Attention Head Intervention
  Jordan Nikolai Pettyjohn, Nathaniel C Hudson, Mansi Sakarvadia, Aswathy Ajith, Kyle Chard

• Can One Token Make All the Difference? Forking Paths in Autoregressive Text Generation
  Eric J Bigelow, Ari Holtzman, Hidenori Tanaka, Tomer Ullman

• Exploring the Recall of Language Models: Case Study on Molecules
  Knarik Mheryan, Hasmik Mnatsakanyan, Philipp Guevorguian, Hrant Khachatrian

• How do LLMs deal with Syntactic Conflicts in In-context-learning ?
  Nahyun Kim

• Do LLMs Use Language Like Humans? Toward A Linguistics Analysis for Understanding LLMs Behavior
  Xinyu Zhou, Delong Chen, Samuel Cahyawijaya, Xufeng Duan, Zhenguang Cai