Finding Transformer Circuits with Edge Pruning
Adithya Bhaskar, Alexander Wettig, Dan Friedman, and Danqi Chen
arXiv 2024
A faster and more precise circuit-finding method that also scales to multi-billion parameter models.
Please see my resume for a more up-to-date list.
Adithya Bhaskar, Alexander Wettig, Dan Friedman, and Danqi Chen
arXiv 2024
A faster and more precise circuit-finding method that also scales to multi-billion parameter models.
Adithya Bhaskar, Dan Friedman, and Danqi Chen
ACL 2024 (Main)
Structured pruning reveals surprising insights about how Pretrained LMs generalize.
Tianyu Gao, Zirui Wang, Adithya Bhaskar, and Danqi Chen
arXiv 2024
Multimodal Language Models can’t read well. We introduce a novel patch-and-text loss to remedy that.
Adithya Bhaskar
EMNLP 2023 (Main)
Current Text-to-SQL conversion systems fall flat on their face when faced with ambiguity. We demonstrate this by introducing a new benchmark (AmbiQT), then propose a novel method improving coverage by up to 2.5x.
Adithya Bhaskar, Alex Fabbri and Greg Durrett
ACL 2023 (Findings)
Novel evaluation metrics for summarization in the GPT-3.5 era.
Richeek Das, Aaron Jerry Ninan, Adithya Bhaskar and Ajit Rajwade
Signal Processing, Vol. 214, January 2024
Performance bounds for Group Testing of, e.g., COVID-19.