2026
Sufficiency is Relative: Evaluating LLM Explanations under Model-Induced Input Distributions
ICML 2026poster
Large language models (LLMs) are increasingly deployed in high-stakes domains, where free-text explanations such as chain-of-thought and post-hoc rationales are used to justify model outputs. Yet it remains unclear whether these explanations are _sufficient_, i.e., if they contain enough information…