Siddharth Suresh

Hi!

I am Siddharth, a third-year graduate student in the Psychology Department and the Computer Science Department at the University of Wisconsin - Madison. I am interested in the intersection of Cognitive Science and Artificial Intelligence. My research interests include using computational modelling techniques like Deep Learning and Machine Learning to understand our brains reprsent semantic information. I also use techniques from Cognitive Science to understand semantic representations in AI models.

Research Interests

Cognitive Science
Representation Learning
Language Models - LLM Agents, SFT, Reasoning, Evaluation, Task Vectors, etc.
Multimodal Learning

Latest News

March 2025: 'AI-enhanced semantic feature norms for 786 concepts.' accepted as an Oral at the ICLR 2025 Birectional Human-AI Alignment Workshop.
March 2025: Received the Schwartz fellowship ($5000) to support research on understanding how task objective shapes representations.
December 2024: I will be co-organising the ICLR 2025 Workshop on Representational Alignment (Re-Align).
November 2024: Accepted offer as a ML Research Intern at Netflix for Summer 2025 to work on LLM agents. Sunny California, here I come!
September 2024: Paper "Probing LLM World Model: Enhancing Guesstimation with Wisdom of Crowds Decoding" accepted to NeurIPS 2024 Workshop on Behavioral Machine Learning
September 2024: Paper "Humor in AI: Massive Scale Crowd-Sourced Preferences and Benchmarks for Cartoon Captioning" accepted to NeurIPS 2024 datasets and benchmarks track
June 2024: Started as an Applied Scientist Intern at Amazon AGI working on generative recommendation systems using LLMs
April 2024: Paper "Learning interactions to boost human creativity with bandits and GPT-4" accepted to CogSci 2024 in Rotterdam
March 2024: Gave an invited talk on "Semantic reprresentations in minds and machines" at PDP seminar at Princeton University

March 2024: Paper "Siimulating Opinion Dynamics with Networks of LLM-based Agents"" accepted to NAACL 2024

February 2024: Two papers accepted at ICLR 2024 workshop on LLM Agents: "The Wisdom of Partisan Crowds" and "Simulating Opinion Dynamics with Networks of LLM-based Agents"

February 2024: Two papers accepted at ICLR 2024 Workshop on Representational Alignment: "Categories vs Semantic Features: What shapes the similarities people discern in photographs of objects?" and "The role of shared labels and experiences in representational alignment" with K. Mukherjee

December 2023: Received departmental travel award of $1500 to present research at EMNLP 2023 in Singapore

November 2023: Gave a guest lecture on using version control for research projects at AI + Society seminar (Wisconsin Institute of Discovery)

September 2023: Gave an invited talk on "Semantic representations in minds and machines" at the MRC Cognition and Brain Sciences Unit, Cambridge, UK

September 2023: Paper "Conceptual structure coheres in human cognition but not in large language models" accepted to the main track at EMNLP 2023

July 2023: Presented poster on "Behavioral estimates of conceptual structure" at CogSci 2023 in Sydney, Australia

June 2023: Received Menzies award of $500 to conduct research using Large Language Models

May 2023: Presented talk on "Can deep convolutional networks explain the semantic structure that humans see in photographs?" at Vision Science Society conference in St. Pete's Beach, Florida