Home  ·  Publications  ·  Background  ·  Awards  ·  Book  ·  Talks

Bodhisattwa Prasad Majumder
bodhisattwam[at]allenai.org
Office @ Northlake Commons
Allen Institute for AI

I am a Senior Research Scientist at Allen Institute for AI. I lead research on autonomous data-driven discovery for Asta, Ai2's agentic ecosystem for scientific research.

I received my Ph.D. in Computer Science from UC San Diego, advised by Julian McAuley. I was fortunate to receive UCSD CSE Doctoral Dissertation Award (2023), UCSD CSE Award for Research Excellence (2022) Adobe Research Fellowship (2022), Qualcomm Innovation Fellowship (2020), and led UCSD (Team Bernard) in Amazon Alexa Prize, 2019.


Google Scholar




Github  |  LinkedIn





I value accessible scientific communication. I worked as a technical editor for The Batch, a weekly newsletter by deeplearning.ai. I also co-authored a best-selling book on Natural Language Processing with O'Reilly Media.


Research

My research goal is to develop communicative reasoners that can learn, adapt, and reason by interacting with the world and produce effective, explainable, and equitable outcomes. I apply my research to advance the frontier of scientific domains such as cancer research, neuroscience, and computer science.

SCIENTIFIC DISCOVERY

I introduced the use of LLMs in data-driven discovery, a heuristic search problem, where given a set of observations, we must uncover the hypotheses, laws, and theories that govern its data-generating process. I combine LLM reasoning, classical search, and probabilistic inference to tackle the core problems of autonomous long-horizon scientific discovery.

LANGUAGE AGENTS

Agents are the cornerstone of daily workstacks. I focus on in-context policy optimization to enable continual learning and rapid task adaptation in language agents. My work focuses on cross-environment and cross-task generalization in a zero-shot setup, by developing reflective, non-parametric memory-driven architectures for partially observable worlds.

SOCIAL INTELLIGENCE

My discovery and language agents work extends to computationally modeling human behavior and personalized preferences with and for LLMs.

HUMAN BEHAVIOR WITH AI

My computational work complements my interest in decoding human behavior around emerging technologies such as AI using tools from behavioral science.

See full body of my work in Publications.
Media

My work humbly finds its way into the press. Here is a selection.

  • The AI Economy | Ai2's AutoDiscovery Uses AI to Find Hidden Hypotheses in Big Data
  • SiliconAngle | Ai2 Introduces AutoDiscovery, Automated Scientific Discovery AI System
  • TechTarget | CAIA Turns to Federated Learning, AI Agents for Cancer Research
  • BigDataWire | Ai2's DataVoyager Lets Scientists Talk to Their Data
  • Business Wire | HPCWire | Yahoo Finance | Ai2 Launches Asta DataVoyager: Data-driven Discovery and Analysis for Science
  • The AI Economy | Ai2's Asta DataVoyager Simplifies Scientific Data Analysis
  • Financial Content | AI Unleashed: Fred Hutch Leads Groundbreaking Alliance to Revolutionize Cancer Research
  • TechStrong | Ai2 Debuts DataVoyager Tool to Democratize Scientific Data Analysis
  • GeekWire | Cancer AI Alliance Says New Tech Platform Will Speed Breakthroughs with Novel Privacy Approach
  • Newsweek | Cancer Researchers Find a Way Around AI's Biggest Bottleneck: Data Sharing
  • KUOW NPR | New AI Platform Led by Fred Hutch Aims to Accelerate Cancer Breakthroughs
  • MIT Technology Review | Google's New Project Astra Could Be Generative AI's Killer App
  • UCSD CSE News | PhD Student Receives Prestigious Adobe Research Fellowship to Improve Interactive Machines
  • UCSD Jacobs School of Engineering | Featured for Qualcomm Innovation Fellowship
  • Google AI Blog | Extracting Structured Data from Templatic Documents



Thanks to Jon Barron for this nice template!
Gorgeous Geisel Library cover art from here.