Sameep Mehta


2023

pdf bib
CFL: Causally Fair Language Models Through Token-level Attribute Controlled Generation
Rahul Madhavan | Rishabh Garg | Kahini Wadhawan | Sameep Mehta
Findings of the Association for Computational Linguistics: ACL 2023

We propose a method to control the attributes of Language Models (LMs) for the text generation task using Causal Average Treatment Effect (ATE) scores and counterfactual augmentation. We explore this method, in the context of LM detoxification, and propose the Causally Fair Language (CFL) architecture for detoxifying pre-trained LMs in a plug-and-play manner. Our architecture is based on a Structural Causal Model (SCM) that is mathematically transparent and computationally efficient as compared with many existing detoxification techniques. We also propose several new metrics that aim to better understand the behaviour of LMs in the context of toxic text generation. Further, we achieve state of the art performance for toxic degeneration, which are computed using Real Toxicity Prompts. Our experiments show that CFL achieves such a detoxification without much impact on the model perplexity. We also show that CFL mitigates the unintended bias problem through experiments on the BOLD dataset.

2013

pdf bib
NLP for uncertain data at scale
Sameep Mehta | L. V. Subramaniam
NAACL HLT 2013 Tutorial Abstracts

pdf bib
An Empirical Assessment of Contemporary Online Media in Ad-Hoc Corpus Creation for Social Events
Kanika Narang | Seema Nagar | Sameep Mehta | L V Subramaniam | Kuntal Dey
Proceedings of the Sixth International Joint Conference on Natural Language Processing