Removing Spurious Correlation from Neural Network Interpretations
The existing algorithms for identification of neurons responsible for undesired and harmful behaviors do not consider the effects of confounders such as topic of the conversation. In this work, we show that confounders can create spurious correlations and propose a new causal mediation approach that controls the impact of the topic. In experiments with two large language models, we study the localization hypothesis and show that adjusting for the effect of conversation topic, toxicity becomes less localized.
Code (0)
등록된 구현이 없습니다.
Similar Papers 제목 키워드 기반
Influence Tuning: Demoting Spurious Correlations via Instance Attribution and Instance-Driven Updates
Among the most critical limitations of deep learning NLP models are their lack of interpretability, and their reliance on spurious correlations. Prior work proposed various approaches to interpreting the black-box models…
Seeing What's Not There: Spurious Correlation in Multimodal LLMs
Unimodal vision models are known to rely on spurious correlations, but it remains unclear to what extent Multimodal Large Language Models (MLLMs) exhibit similar biases despite language supervision. In this paper, we inv…
HallucinationObjectObject HallucinationObject RecognitionRemoving Spurious Concepts from Neural Network Representations via Joint Subspace Estimation
Out-of-distribution generalization in neural networks is often hampered by spurious correlations. A common strategy is to mitigate this by removing spurious concepts from the neural network representation of the data. Ex…
Out-of-Distribution GeneralizationIdentifying and Disentangling Spurious Features in Pretrained Image Representations
Neural networks employ spurious correlations in their predictions, resulting in decreased performance when these correlations do not hold. Recent works suggest fixing pretrained representations and training a classificat…
Reducing Spurious Correlations for Answer Selection by Feature Decorrelation and Language Debiasing
Deep neural models have become the mainstream in answer selection, yielding state-of-the-art performance. However, these models tend to rely on spurious correlations between prediction labels and input features, which in…
Answer SelectionContrastive Learning