CoBRA
CoBRA: Programming Cognitive Bias in Social Agents Using Classic Social Science Experiments
Calibrate and steer agent behavior across input, activation, and parameter spaces, grounded in classic social science experiments.
Research
HumanStudy-Hub
HumanStudy-Bench: Towards AI Agent Design for Participant Simulation
Turn human studies into reusable datasets and runnable experiments for AI agents.
Project
CogMir
Exploring Prosocial Irrationality for LLM Agents: A Social Cognition View
Reframe and harness LLM hallucinations through the lens of human cognitive biases and prosocial irrationality.
Paper
CoThinker
United Minds or Isolated Agents? Exploring Coordination of LLMs under Cognitive Load Theory
Overcome cognitive overload through distributed reasoning and shared memory.
Paper