Social World Models

Simons Institute for the Theory of Computing · Beginner ·👁️ Computer Vision ·3w ago

About this lesson

Paul Liang (MIT) https://simons.berkeley.edu/talks/paul-liang-mit-2026-06-12 Topics in Intelligence: World Models and Social Reasoning Humans navigate complex social environments by continuously predicting the behavior, intentions, and beliefs of others. Inspired by this ability, Social World Models extend existing physical world models to the social domain, capturing not only physical dynamics but also the latent mental states and strategies of interacting agents. This talk will explore how these models can be learned, represented, and leveraged to build towards socially intelligent AI. We will present our recent work in multimodal modeling of social behaviors, social reasoning benchmarks and models, and long-horizon self-evolving social agents.

Original Description

Paul Liang (MIT) https://simons.berkeley.edu/talks/paul-liang-mit-2026-06-12 Topics in Intelligence: World Models and Social Reasoning Humans navigate complex social environments by continuously predicting the behavior, intentions, and beliefs of others. Inspired by this ability, Social World Models extend existing physical world models to the social domain, capturing not only physical dynamics but also the latent mental states and strategies of interacting agents. This talk will explore how these models can be learned, represented, and leveraged to build towards socially intelligent AI. We will present our recent work in multimodal modeling of social behaviors, social reasoning benchmarks and models, and long-horizon self-evolving social agents.

Watch on YouTube ↗ (saves to browser)

Sign in to unlock AI tutor explanation · ⚡30

Related Reads

PANet Paper Walkthrough: When Feature Pyramids Go Bottom-Up

Learn how PANet's bottom-up feature pyramid approach improves feature extraction by shortening the path between low-level and high-level features

Towards Data Science

CCTV Action Recognition: Comprehensive Fine-Tuning & Real-Time Deployment Guide

Learn to fine-tune and deploy a hybrid Deep Learning model for CCTV action recognition using MobileNetV2 and Python

Medium · Python

I built a background remover that keeps the fine hair edges

Learn how to build a background remover that preserves fine hair edges, a challenging task in image processing

Dev.to · KunStudio

I Built a Python Package to Solve My Own CV Frustration — 7K Downloads in a Week

Learn how to create a Python package to simplify computer vision pipelines and achieve 7K downloads in a week

Medium · Machine Learning

Marketing management for ugc net| Important topics of marketing management ugc net commerce dec 2023

Bhoomi Learning Centre~Dr. Muskan