paper-with-me

Papers

Causal Mediation Analysis for Interpreting Neural NLP: The Case of Gender Bias

2020-04-26 · Jesse Vig, Sebastian Gehrmann, Yonatan Belinkov, Sharon Qian, Daniel Nevo, Simas Sakenis, Jason Huang, Yaron Singer, Stuart Shieber

Common methods for interpreting neural models in natural language processing typically examine either their structure or their behavior, but not both. We propose a methodology grounded in the theory of causal mediation analysis for interpreting which parts of a model are causally implicated in its behavior. It enables us to analyze the mechanisms by which information flows from input to output through various model components, known as mediators. We apply this methodology to analyze gender bias in pre-trained Transformer language models. We study the role of individual neurons and attention heads in mediating gender bias across three datasets designed to gauge a model's sensitivity to gender bias. Our mediation analysis reveals that gender bias effects are (i) sparse, concentrated in a small part of the network; (ii) synergistic, amplified or repressed by different components; and (iii) decomposable into effects flowing directly from the input and indirectly through the mediators.

📄 PDF Abstract BibTeX arXiv:2004.12265

Code (1)

sebastianGehrmann/CausalMediationAnalysis 공식 구현 pytorch

Methods 이 논문이 사용한 방법론

Linear Layer A Linear Layer is a projection $\mathbf{XW + b}$.
Absolute Position Encodings Absolute Position Encodings are a type of position embeddings for [Transformer-based models] where positional encodings are…
Position-Wise Feed-Forward Layer 설명 없음
Residual Connection 설명 없음
BPE Byte Pair Encoding, or BPE, is a subword segmentation algorithm that encodes rare and unknown words as sequences of subword units. The intuition is that various word…
Dense Connections Dense Connections, or Fully Connected Connections, are a type of layer in a deep neural network that use a linear operation where every input is connected to every output…
Label Smoothing Label Smoothing is a regularization technique that introduces noise for the labels. This accounts for the fact that datasets may have mistakes in them, so maximizing the…
ReLU How Do I Communicate to Expedia? How Do I Communicate to Expedia? – Call ☎️ +1-(888) 829 (0881) or +1-805-330-4056 or +1-805-330-4056 for Live Support & Special Travel…

Similar Papers 제목 키워드 기반

Investigating Gender Bias in Language Models Using Causal Mediation Analysis

2020-12-01 · NeurIPS 2020 12 · Jesse Vig, Sebastian Gehrmann, Yonatan Belinkov, Sharon Qian 외

Many interpretation methods for neural models in natural language processing investigate how information is encoded inside hidden representations. However, these methods can only measure whether the information exists, n…

Interpreting Bias in Large Language Models: A Feature-Based Approach

2024-06-18 · Nirmalendu Prakash, Lee Ka Wei Roy

Large Language Models (LLMs) such as Mistral and LLaMA have showcased remarkable performance across various natural language processing (NLP) tasks. Despite their success, these models inherit social biases from the dive…

counterfactual

What Changed? Investigating Debiasing Methods using Causal Mediation Analysis

2022-06-01 · NAACL (GeBNLP) 2022 7 · Sullam Jeoung, Jana Diesner

Previous work has examined how debiasing language models affect downstream tasks, specifically, how debiasing techniques influence task performance and whether debiased models also make impartial predictions in downstrea…

Scaling Causal Mediation for Complex Systems: A Framework for Root Cause Analysis

2025-12-16 · Alessandro Casadei, Sreyoshi Bhaduri, Rohit Malshe, Pavan Mullapudi 외 arxiv

Modern operational systems ranging from logistics and cloud infrastructure to industrial IoT, are governed by complex, interdependent processes. Understanding how interventions propagate through such systems requires cau…

Causal Inference

Total-effect Test May Erroneously Reject So-called "Full" or "Complete" Mediation

2023-09-16 · TingXuan Han, Luxi Zhang, Xinshu Zhao, Ke Deng

The procedure for establishing mediation, i.e., determining that an independent variable X affects a dependent variable Y through some mediator M, has been under debate. The classic causal steps require that a "total eff…

Mathematical Proofs