paper-with-me

홈 › Papers

Invariant Rationalization

2020-03-22 · ICML 2020 1 · Shiyu Chang, Yang Zhang, Mo Yu, Tommi S. Jaakkola

Selective rationalization improves neural network interpretability by identifying a small subset of input features -- the rationale -- that best explains or supports the prediction. A typical rationalization criterion, i.e. maximum mutual information (MMI), finds the rationale that maximizes the prediction performance based only on the rationale. However, MMI can be problematic because it picks up spurious correlations between the input features and the output. Instead, we introduce a game-theoretic invariant rationalization criterion where the rationales are constrained to enable the same predictor to be optimal across different environments. We show both theoretically and empirically that the proposed rationales can rule out spurious correlations, generalize better to different test scenarios, and align better with human judgments. Our data and code are available.

📄 PDF Abstract BibTeX arXiv:2003.09772

Code (1)

code-terminator/invariant_rationalization 공식 구현 tf

Methods 이 논문이 사용한 방법론

Interpretability 설명 없음

Similar Papers 제목 키워드 기반

Fine-grained Graph Rationalization

2023-12-13 · Zhe Xu, Menghai Pan, Yuzhong Chen, Huiyuan Chen 외

Rationale discovery is defined as finding a subset of the input data that maximally supports the prediction of downstream tasks. In the context of graph machine learning, graph rationale is defined to locate the critical…

Prediction

Mitigating Biases in Toxic Language Detection through Invariant Rationalization

2021-06-14 · ACL (WOAH) 2021 8 · Yung-Sung Chuang, Mingye Gao, Hongyin Luo, James Glass 외

Automatic detection of toxic language plays an essential role in protecting social media users, especially minority groups, from verbal abuse. However, biases toward some attributes, including gender, race, and dialect, …

Natural Language Understanding

Cooperative Classification and Rationalization for Graph Generalization

2024-03-10 · Linan Yue, Qi Liu, Ye Liu, Weibo Gao 외

Graph Neural Networks (GNNs) have achieved impressive results in graph classification tasks, but they struggle to generalize effectively when faced with out-of-distribution (OOD) data. Several approaches have been propos…

ClassificationGraph ClassificationKnowledge Distillation

Faithful or Fabricated? A Causal Framework for Rationalization Bias in LLM Judges

2026-05-13 · Riya Tapwal, Abhishek Kumar, Carsten Maple arxiv

Large language models (LLMs) are increasingly used as automatic judges for summarization and dialogue evaluation. Prior work has documented biases such as position, verbosity, and style preferences, but largely focuses o…

Dialogue Evaluation

Rationalization: A Neural Machine Translation Approach to Generating Natural Language Explanations

2017-02-25 · Upol Ehsan, Brent Harrison, Larry Chan, Mark O. Riedl

We introduce AI rationalization, an approach for generating explanations of autonomous system behavior as if a human had performed the behavior. We describe a rationalization technique that uses neural machine translatio…

Explanation GenerationMachine TranslationTranslation