publications

2024

1. NeurIPS 2024

   LLM Targeted Underperformance Disproportionately Impacts Vulnerable Users

   Elinor Poole-Dayan, Deb Roy, and Jad Kabbara

   2024

   arXiv
2. EMNLP 2024

   On the Relationship between Truth and Political Bias in Language Models

   Suyash Fulay, William Brannon, Shrestha Mohanty, Cassandra Overney, Elinor Poole-Dayan, Deb Roy, and Jad Kabbara

   In Proceedings of the 2024 Conference on Empirical Methods in Natural Language Processing, Nov 2024

   Abs arXiv PDF

   Language model alignment research often attempts to ensure that models are not only helpful and harmless, but also truthful and unbiased. However, optimizing these objectives simultaneously can obscure how improving one aspect might impact the others. In this work, we focus on analyzing the relationship between two concepts essential in both language model alignment and political science: truthfulness and political bias. We train reward models on various popular truthfulness datasets and subsequently evaluate their political bias. Our findings reveal that optimizing reward models for truthfulness on these datasets tends to result in a left-leaning political bias. We also find that existing open-source reward models (i.e., those trained on standard human preference datasets) already show a similar bias and that the bias is larger for larger models. These results raise important questions about the datasets used to represent truthfulness, potential limitations of aligning models to be both truthful and politically unbiased, and what language models capture about the relationship between truth and politics.

2023

1. NeurIPS 2023

   Are Diffusion Models Vision-And-Language Reasoners?

   Benno Krojer, Elinor Poole-Dayan, Vikram Voleti, Chris Pal, and Siva Reddy

   In Advances in Neural Information Processing Systems, Nov 2023

   Abs arXiv PDF

   Text-conditioned image generation models have recently shown immense qualitative success using denoising diffusion processes. However, unlike discriminative vision-and-language models, it is a non-trivial task to subject these diffusion-based generative models to automatic fine-grained quantitative evaluation of high-level phenomena such as compositionality.Towards this goal, we perform two innovations. First, we transform diffusion-based models (in our case, Stable Diffusion) for any image-text matching (ITM) task using a novel method called DiffusionITM.Second, we introduce the Generative-Discriminative Evaluation Benchmark (GDBench) benchmark with 7 complex vision-and-language tasks, bias evaluation and detailed analysis.We find that Stable Diffusion + DiffusionITM is competitive on many tasks and outperforms CLIP on compositional tasks like like CLEVR and Winoground.We further boost its compositional performance with a transfer setup by fine-tuning on MS-COCO while retaining generative capabilities. We also measure the stereotypical bias in diffusion models, and find that Stable Diffusion 2.1 is, for the most part, less biased than Stable Diffusion 1.5.Overall, our results point in an exciting direction bringing discriminative and generative model evaluation closer. We will release code and benchmark setup soon.

2022

1. ACL 2022

   An Empirical Survey of the Effectiveness of Debiasing Techniques for Pre-trained Language Models

   Nicholas Meade, Elinor Poole-Dayan, and Siva Reddy

   In Proceedings of the 60th Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers), May 2022

   Abs DOI arXiv PDF

   Recent work has shown pre-trained language models capture social biases from the large amounts of text they are trained on. This has attracted attention to developing techniques that mitigate such biases. In this work, we perform an empirical survey of five recently proposed bias mitigation techniques: Counterfactual Data Augmentation (CDA), Dropout, Iterative Nullspace Projection, Self-Debias, and SentenceDebias. We quantify the effectiveness of each technique using three intrinsic bias benchmarks while also measuring the impact of these techniques on a model’s language modeling ability, as well as its performance on downstream NLU tasks. We experimentally find that: (1) Self-Debias is the strongest debiasing technique, obtaining improved scores on all bias benchmarks; (2) Current debiasing techniques perform less consistently when mitigating non-gender biases; And (3) improvements on bias benchmarks such as StereoSet and CrowS-Pairs by using debiasing strategies are often accompanied by a decrease in language modeling ability, making it difficult to determine whether the bias mitigation was effective.