Related papers: Auditing Stance Asymmetry in Generative Explanatio…

Sparse autoencoders (SAEs) are a popular technique for interpreting language model activations, and there is extensive recent work on improving SAE effectiveness. However, most prior work evaluates progress using unsupervised proxy metrics…

Concept-driven counterfactuals explain decisions of classifiers by altering the model predictions through semantic changes. In this paper, we present a novel approach that leverages cross-modal decompositionality and image-specific concepts…

Word embeddings are the standard model for semantic and syntactic representations of words. Unfortunately, these models have been shown to exhibit undesirable word associations resulting from gender, racial, and religious biases. Existing…

Alignment with high-resource standard languages is often assumed to aid the modeling of related low-resource varieties. We challenge this assumption by demonstrating that excessive representational entanglement with a dominant variety, such…

To recognize and mitigate the harms of generative AI systems, it is crucial to consider whether and how different societal groups are represented by these systems. A critical gap emerges when naively measuring or improving who is…

Targeted syntactic evaluations of language models ask whether models show stable preferences for syntactically acceptable content over minimal-pair unacceptable inputs. Most targeted syntactic evaluation datasets ask models to make these…

It is unclear whether, how and where large pre-trained language models capture subtle linguistic traits like ambiguity, grammaticality and sentence complexity. We present results of automatic classification of these traits and compare their…

While Large Language Models achieve state-of-the-art results across a wide range of NLP tasks, they remain prone to systematic biases. Among these, gender bias is particularly salient in MT, due to systematic differences across languages in…

The rapid evolution of Large Language Models (LLMs) has transformed natural language processing but raises critical concerns about biases inherent in their deployment and use across diverse linguistic and sociocultural contexts. This paper…

Large language models (LLMs) have achieved remarkable progress, yet their internal mechanisms remain largely opaque, posing a significant challenge to their safe and reliable deployment. Sparse autoencoders (SAEs) have emerged as a…

Although Large Language Models (LLMs) demonstrate remarkable reasoning capabilities, inherent social biases often cascade throughout the Chain-of-Thought (CoT) process, leading to continuous "Bias Propagation". Existing debiasing methods…

Automatically evaluating the quality of dialogue responses for unstructured domains is a challenging problem. ADEM(Lowe et al. 2017) formulated the automatic evaluation of dialogue systems as a learning problem and showed that such a model…

Human judgments of word similarity have been a popular method of evaluating the quality of word embedding. But it fails to measure the geometry properties such as asymmetry. For example, it is more natural to say "Ellipses are like Circles"…

Vision-language models can encode societal biases and stereotypes, but there are challenges to measuring and mitigating these multimodal harms due to lacking measurement robustness and feature degradation. To address these challenges, we…

There has been considerable interest in using decomposition methods in epidemiology (mediation analysis) and economics (Oaxaca-Blinder decomposition) to understand how health disparities arise and how they might change upon intervention. It…

Multilingual studies of social bias in open-ended LLM generation remain limited: most existing benchmarks are English-centric, template-based, or restricted to recognizing pre-specified stereotypes. We introduce StereoTales, a multilingual…

We present a supervised learning framework of training generative models for density estimation. Generative models, including generative adversarial networks, normalizing flows, variational auto-encoders, are usually considered as…

Recent research demonstrates that word embeddings, trained on the human-generated corpus, have strong gender biases in embedding spaces, and these biases can result in the discriminative results from the various downstream tasks. Whereas…

In attempts to "explain" predictions of machine learning models, researchers have proposed hundreds of techniques for attributing predictions to features that are deemed important. While these attributions are often claimed to hold the…

Interpretability can be implemented to understand decisions taken by (black box) models, such as neural machine translation (NMT) or large language models (LLMs). Yet, research in this area has been limited in relation to a manifested…