Ramakanth Pasunuru

I am a Research Scientist at FAIR, Meta. I received my Ph.D. in 2021 from the Computer Science Department, UNC Chapel Hill, where I was advised by Prof. Mohit Bansal. I received my B.Tech and M.Tech from IIT Kharagpur in 2015.

Google Scholar / GitHub / Twitter / ACL Anthology

Research

My research focuses on the intersection of large language models and computer vision. I work on developing and scaling multimodal foundation models such as Chameleon, as well as language-model pre-training, including Byte Latent Transformer, and post-training methods for stronger reasoning and alignment.

Selected Papers

For the complete and most current list, please see my Google Scholar page.

Figure from the HoneyBee paper

HoneyBee: Data Recipes for Vision-Language Reasoners

Hritik Bansal, Devandra Sachan, Kai-Wei Chang, Aditya Grover, Gargi Ghosh, Wen-tau Yih, Ramakanth Pasunuru

Proceedings of CVPR 2026

Figure from the Byte Latent Transformer paper

Byte Latent Transformer: Patches Scale Better Than Tokens

Artidoro Pagnoni, Ramakanth Pasunuru, Pedro Rodriguez, et al.

Proceedings of ACL 2025, Outstanding Paper Award

Figure from the Chain-of-Abstraction paper

Efficient Tool Use with Chain-of-Abstraction Reasoning

Silin Gao, Jane Dwivedi-Yu, Ping Yu, Xiaoqing Ellen Tan, Ramakanth Pasunuru, et al.

Proceedings of COLING 2025

Figure from the Chameleon paper

Chameleon: Mixed-Modal Early-Fusion Foundation Models

Srinivasan Iyer*, Ramakanth Pasunuru*, Bernie Huang*, Armen Aghajanyan*, et al. (*Joint first authors)

Preprint, 2024

Figure from the ART of LLM Refinement paper

The ART of LLM Refinement: Ask, Refine, and Trust

Kumar Shridhar, Koustuv Sinha, Andrew Cohen, Tianlu Wang, Ping Yu, Ram Pasunuru, et al.

NAACL 2024

Figure from the Augmented Language Models survey

Augmented Language Models: a Survey

Grégoire Mialon, Roberto Dessì, Maria Lomeli, ... Ram Pasunuru, ... Thomas Scialom

Proceedings of TMLR 2023

Figure from the CRYSTAL paper

CRYSTAL: Introspective Reasoners Reinforced with Self-Feedback

Jiacheng Liu, Ramakanth Pasunuru, Hannaneh Hajishirzi, Yejin Choi, and Asli Celikyilmaz

EMNLP 2023

Figure from the Scaling Autoregressive Multi-Modal Models paper

Scaling Autoregressive Multi-Modal Models: Pretraining and Instruction Tuning

Lili Yu*, Bowen Shi*, Ramakanth Pasunuru*, et al. (*Joint first authors)

Preprint, 2023

Figure from the OPT-IML paper

OPT-IML: Scaling Language Model Instruction Meta Learning through the Lens of Generalization

Srinivasan Iyer*, Xi Victoria Lin*, Ramakanth Pasunuru*, et al. (*Joint first authors)

Preprint, 2022

Figure from the multilingual few-shot learning paper

Few-shot Learning with Multilingual Generative Language Models

Xi Victoria Lin, ... Ramakanth Pasunuru, ... Xian Li

Proceedings of EMNLP 2022

Figure from the architecture search paper

Continual and Multi-Task Architecture Search

Ramakanth Pasunuru, Mohit Bansal

Proceedings of ACL 2019

Figure from the multi-reward summarization paper

Multi-Reward Reinforced Summarization with Saliency and Entailment

Ramakanth Pasunuru, Mohit Bansal

Proceedings of NAACL 2018

Figure from the video captioning paper

Multi-Task Video Captioning with Video and Entailment Generation

Ramakanth Pasunuru, Mohit Bansal

Proceedings of ACL 2017, Outstanding Paper Award