Proposes SCSuff metric for evaluating LLM explanation sufficiency via model-generated alternative inputs, showing explanations are typically insufficient and predictable from hidden states.
Do LLMs overcome shortcut learning? An evaluation of shortcut challenges in large language models
4 Pith papers cite this work. Polarity classification is still indexing.
4
Pith papers citing it
verdicts
UNVERDICTED 4representative citing papers
NeuroFlake integrates discriminative token mining into LLMs to classify flaky tests, raising F1-score to 69.34% on FlakeBench while showing greater robustness to semantic-preserving perturbations than prior methods.
Evolutionary game theory shows gradient descent and stochastic gradient descent drive neural networks to distinct stable states favoring shortcut or core subnetworks, with data and optimization noise shaping shortcut bias formation.
Empirical tests of VLM-CBMs show VLM supervision differs from expert annotations depending on task and that concept accuracy correlates weakly with quality metrics.
citing papers explorer
-
What LLMs explain is not what they believe: Evaluating explanation sufficiency under models' own input beliefs
Proposes SCSuff metric for evaluating LLM explanation sufficiency via model-generated alternative inputs, showing explanations are typically insufficient and predictable from hidden states.
-
NeuroFlake: A Neuro-Symbolic LLM Framework for Flaky Test Classification
NeuroFlake integrates discriminative token mining into LLMs to classify flaky tests, raising F1-score to 69.34% on FlakeBench while showing greater robustness to semantic-preserving perturbations than prior methods.
-
Deciphering Shortcut Learning from an Evolutionary Game Theory Perspective
Evolutionary game theory shows gradient descent and stochastic gradient descent drive neural networks to distinct stable states favoring shortcut or core subnetworks, with data and optimization noise shaping shortcut bias formation.
-
If Concept Bottlenecks are the Question, are Foundation Models the Answer?
Empirical tests of VLM-CBMs show VLM supervision differs from expert annotations depending on task and that concept accuracy correlates weakly with quality metrics.