Josh Barua

Keegan Quigley, Miriam Cha, Josh Barua et al.

Vision-language pretraining has been shown to produce high-quality visual encoders which transfer efficiently to downstream computer vision tasks. Contrastive learning approaches have increasingly been adopted for medical vision language pretraining (MVLP), yet recent developments in generative AI offer new modeling alternatives. This paper introduces RadTex, a CNN-encoder transformer-decoder architecture optimized for radiology. We explore bidirectional captioning as an alternative MVLP strategy and demonstrate that RadTex's captioning pretraining is competitive with established contrastive methods, achieving a CheXpert macro-AUC of 89.4%. Additionally, RadTex's lightweight text decoder not only generates clinically relevant radiology reports (macro-F1 score of 0.349), but also provides targeted, interactive responses, highlighting the utility of bidirectional captioning in advancing medical image analysis.

Josh Barua, Sanjay Subramanian, Kayo Yin et al.

In translation, a concept represented by a single word in a source language can have multiple variations in a target language. The task of lexical selection requires using context to identify which variation is most appropriate for a source text. We work with native speakers of nine languages to create DTAiLS, a dataset of 1,377 sentence pairs that exhibit cross-lingual concept variation when translating from English. We evaluate recent LLMs and neural machine translation systems on DTAiLS, with the best-performing model, GPT-4, achieving from 67 to 85% accuracy across languages. Finally, we use language models to generate English rules describing target-language concept variations. Providing weaker models with high-quality lexical rules improves accuracy substantially, in some cases reaching or outperforming GPT-4.

Josh Barua, Seun Eisape, Kayo Yin et al.

While large reasoning models have shown remarkable ability to generate long chains-of-thought (CoTs) in English, we still lack understanding of how these long-form reasoning abilities transfer to the vast majority of the world's languages. In this work, we systematically investigate four key stages of model development--scaling, pretraining, post-training, and inference--to understand how long CoT capabilities extend beyond English. We compare two reasoning settings across nine non-English target languages: En-CoT, where models process target-language inputs, but reason in English; and Target-CoT, where models both process inputs and generate long CoTs in the target language. We find that scaling reasoning model size improves multilingual task performance in En-CoT, but Target-CoT performance lags behind. This gap widens for tasks requiring long, multi-step CoTs such as mathematical reasoning. Shifting to pretraining, we find that adding a specialized reasoning stage enhances En-CoT performance but degrades Target-CoT, whereas broad multilingual pretraining improves both modes simultaneously. Given the scarcity of high-quality reasoning traces in languages other than English, we explore synthetic data curation approaches for post-training. We demonstrate that fine-tuning on reasoning traces automatically translated from gold English traces outperforms fine-tuning on target-language traces distilled from large reasoning models. Finally, we report disparities in inference efficiency between languages and uncover language-specific failure modes in CoTs. We release models, datasets, and code to foster further research.