Selective Brain Damage: Measuring the Disparate Impact of Model Pruning
Neural network pruning techniques have demonstrated it is possible to remove the majority of weights in a network with surprisingly little degradation to top-1 test set accuracy. However, this measure of performance conceals significant differences in how different classes and images are impacted by pruning. We find that certain individual data points, which we term pruning identified exemplars (PIEs), and classes are systematically more impacted by the introduction of sparsity. Removing PIE images from the test-set greatly improves top-1 accuracy for both sparse and non-sparse models. These hard-to-generalize-to images tend to be of lower image quality, mislabelled, entail abstract representations, require fine-grained classification or depict atypical class examples.
Code (0)
등록된 구현이 없습니다.
Tasks
Network PruningMethods 이 논문이 사용한 방법론
Similar Papers 제목 키워드 기반
Local White Matter Architecture Defines Functional Brain Dynamics
Large bundles of myelinated axons, called white matter, anatomically connect disparate brain regions together and compose the structural core of the human connectome. We recently proposed a method of measuring the local …
Functional ConnectivityVariable SelectionFrom Biased Selective Labels to Pseudo-Labels: An Expectation-Maximization Framework for Learning from Biased Decisions
Selective labels occur when label observations are subject to a decision-making process; e.g., diagnoses that depend on the administration of laboratory tests. We study a clinically-inspired selective label problem calle…
Decision MakingDiagnosticRelationship between brain injury criteria and brain strain across different types of head impacts can be different
Multiple brain injury criteria (BIC) are developed to quickly quantify brain injury risks after head impacts. These BIC originated from different types of head impacts (e.g., sports and car crashes) are widely used in ri…
regressionFAIRLABEL: Correcting Bias in Labels
There are several algorithms for measuring fairness of ML models. A fundamental assumption in these approaches is that the ground truth is fair or unbiased. In real-world datasets, however, the ground truth often contain…
FairnessOBCache: Optimal Brain KV Cache Pruning for Efficient Long-Context LLM Inference
Large language models (LLMs) with extended context windows enable powerful applications but impose significant memory overhead, as caching all key-value (KV) states scales linearly with sequence length and batch size. Ex…