Interpretability

Prediction without Preclusion: Recourse Verification with Reachable Sets ICLR 2024 Predictive, scalable and interpretable knowledge tracing on structured domains ICLR 2024 Preemptive Answer “Attacks” on Chain-of-Thought Reasoning ACL 2024 Preference Tuning For Toxicity Mitigation Generalizes Across Languages EMNLP 2024 Pre-trained Language Models Return Distinguishable Probability Distributions to Unfaithfully Hallucinated Texts EMNLP 2024 Pre-trained Large Language Models Use Fourier Features to Compute Addition NeurIPS 2024 Preventing Arbitrarily High Confidence on Far-Away Data in Point-Estimated Discriminative Neural Networks AISTATS 2024 Pride and Prejudice: LLM Amplifies Self-Bias in Self-Refinement ACL 2024 Principal eigenstate classical shadows COLT 2024 Prism: A Framework for Decoupling and Assessing the Capabilities of VLMs NeurIPS 2024 Probabilistic Conceptual Explainers: Trustworthy Conceptual Explanations for Vision Foundation Models ICML 2024 Probabilistic Constrained Reinforcement Learning with Formal Interpretability ICML 2024 Probing Critical Learning Dynamics of PLMs for Hate Speech Detection EACL 2024 Probing Language Models for Pre-training Data Detection ACL 2024 Probing Language Models on Their Knowledge Source EMNLP 2024 Probing Large Language Models from a Human Behavioral Perspective COLING 2024 Probing the Category of Verbal Aspect in Transformer Language Models NAACL 2024 Program-Aided Reasoners (Better) Know What They Know NAACL 2024 Prometheus 2: An Open Source Language Model Specialized in Evaluating Other Language Models EMNLP 2024 Prometheus: Inducing Fine-Grained Evaluation Capability in Language Models ICLR 2024 Promoting Counterfactual Robustness through Diversity AAAI 2024 Prompt-based Generation of Natural Language Explanations of Synthetic Lethality for Cancer Drug Discovery COLING 2024 PromptExplainer: Explaining Language Models through Prompt-based Learning EACL 2024 Prompt Highlighter: Interactive Control for Multi-Modal LLMs CVPR 2024 Prompting Explicit and Implicit Knowledge for Multi-hop Question Answering Based on Human Reading Process COLING 2024