Image de couverture de livre

Do LLM Self-Explanations Help Users Predict Model Behavior? Evaluating Counterfactual Simulatability with Pragmatic Perturbations

Fuente: arXiv

Enregistré dans:

Détails bibliographiques
Auteurs principaux:	Hong, Pingjun, Roth, Benjamin
Format:	Preprint
Publié:	2026
Sujets:	Computation and Language
Accès en ligne:	Acceder al recurso
Tags:	Ajouter un tag Pas de tags, Soyez le premier à ajouter un tag!

Documents similaires

Not All Explanations Simulate Equally: Comparing Verbalized Feature Attributions and Self-Generated Rationales
par: Hong, Pingjun, et autres
Publié: (2026)

Counterfactual Simulatability of LLM Explanations for Generation Tasks
par: Limpijankit, Marvin, et autres
Publié: (2025)

Human Label Variation as Stable Signal: Learning Annotator-Specific Explanation Behavior via Cross-Annotator Preference Optimization
par: Chen, Beiduo, et autres
Publié: (2026)

Agree, Disagree, Explain: Decomposing Human Label Variation in NLI through the Lens of Explanations
par: Hong, Pingjun, et autres
Publié: (2025)

ConSim: Measuring Concept-Based Explanations' Effectiveness with Automated Simulatability
par: Poché, Antonin, et autres
Publié: (2025)

Select or Project? Evaluating Lower-dimensional Vectors for LLM Training Data Explanations
par: Hinterleitner, Lukas, et autres
Publié: (2026)

LiTEx: A Linguistic Taxonomy of Explanations for Understanding Within-Label Variation in Natural Language Inference
par: Hong, Pingjun, et autres
Publié: (2025)

Evaluating Large Language Models for Cross-Lingual Retrieval
par: Zuo, Longfei, et autres
Publié: (2025)

An Evaluation of Explanation Methods for Black-Box Detectors of Machine-Generated Text
par: Schoenegger, Loris, et autres
Publié: (2024)

ALMANACS: A Simulatability Benchmark for Language Model Explainability
par: Mills, Edmund, et autres
Publié: (2023)

Compact Example-Based Explanations for Language Models
par: Schoenegger, Loris, et autres
Publié: (2026)

Prompt-Counterfactual Explanations for Generative AI System Behavior
par: Goethals, Sofie, et autres
Publié: (2026)

LLM-Generated Black-box Explanations Can Be Adversarially Helpful
par: Ajwani, Rohan, et autres
Publié: (2024)

Faithfulness vs. Safety: Evaluating LLM Behavior Under Counterfactual Medical Evidence
par: Mo, Kaijie, et autres
Publié: (2026)

Counterfactual Reasoning with Knowledge Graph Embeddings
par: Zellinger, Lena, et autres
Publié: (2024)

Helpful assistant or fruitful facilitator? Investigating how personas affect language model behavior
par: de Araujo, Pedro Henrique Luz, et autres
Publié: (2024)

DRIV-EX: Counterfactual Explanations for Driving LLMs
par: Cardiel, Amaia, et autres
Publié: (2026)

Evaluating the Reliability of Self-Explanations in Large Language Models
par: Randl, Korbinian, et autres
Publié: (2024)

What if you said that differently?: How Explanation Formats Affect Human Feedback Efficacy and User Perception
par: Malaviya, Chaitanya, et autres
Publié: (2023)

Self-Explanation in Social AI Agents
par: Basappa, Rhea, et autres
Publié: (2025)

Towards Unifying Evaluation of Counterfactual Explanations: Leveraging Large Language Models for Human-Centric Assessments
par: Domnich, Marharyta, et autres
Publié: (2024)

Tuning Language Models for Robust Prediction of Diverse User Behaviors
par: Meng, Fanjin, et autres
Publié: (2025)

Experimental Pragmatics with Machines: Testing LLM Predictions for the Inferences of Plain and Embedded Disjunctions
par: Tsvilodub, Polina, et autres
Publié: (2024)

NeuroFaith: Evaluating LLM Self-Explanation Faithfulness via Internal Representation Alignment
par: Bhan, Milan, et autres
Publié: (2025)

PLEX: Perturbation-free Local Explanations for LLM-Based Text Classification
par: Rahulamathavan, Yogachandran, et autres
Publié: (2025)

Counterfactual LLM-based Framework for Measuring Rhetorical Style
par: Qiu, Jingyi, et autres
Publié: (2025)

Beyond Isolated Behaviors: Hierarchical User Modeling for LLM Personalization
par: Wang, Liang, et autres
Publié: (2026)

Bisimilarity and Simulatability of Processes Parameterized by Join Interactions
par: Grabmayer, Clemens, et autres
Publié: (2025)

When Do "More Contexts" Help with Sarcasm Recognition?
par: Nimase, Ojas, et autres
Publié: (2024)

Counterfactual Graph for Multi-Agent LLM Calibration
par: Huang, Jiatan, et autres
Publié: (2026)

Do language models accommodate their users? A study of linguistic convergence
par: Blevins, Terra, et autres
Publié: (2025)

Characterizing LLM Abstention Behavior in Science QA with Context Perturbations
par: Wen, Bingbing, et autres
Publié: (2024)

Counterfactual Explanations for MITL Violations
par: Finkbeiner, Bernd, et autres
Publié: (2024)

Aligning What LLMs Do and Say: Towards Self-Consistent Explanations
par: Admoni, Sahar, et autres
Publié: (2025)

Natural Language Counterfactual Explanations for Graphs Using Large Language Models
par: Giorgi, Flavio, et autres
Publié: (2024)

Pragmatic Competence Evaluation of Large Language Models for the Korean Language
par: Park, Dojun, et autres
Publié: (2024)

The Illusion of Competence: Evaluating the Effect of Explanations on Users' Mental Models of Visual Question Answering Systems
par: Sieker, Judith, et autres
Publié: (2024)

Safety Training Persists Through Helpfulness Optimization in LLM Agents
par: Plaut, Benjamin
Publié: (2026)

LLM Self-Explanations Fail Semantic Invariance
par: Szeider, Stefan
Publié: (2026)

Do Metrics for Counterfactual Explanations Align with User Perception?
par: Liedeker, Felix, et autres
Publié: (2026)