aiqbench analysis
This research introduces the Social World Model (SWM), a framework that leverages LLMs to predict shifts in social beliefs using temporal patterns from prediction markets like Kalshi and Polymarket. By optimizing evidence lower bounds without requiring explicit human annotations or census data, the model achieves state-of-the-art performance on real-world datasets while providing interpretable insights into social dynamics. The work establishes a new benchmark with over 12k data points, offering a scalable approach for modeling complex socio-political and financial events.
Building Social World Models with Large Language Models
arXiv:2606.11482v1 Announce Type: cross Abstract: Understanding and predicting how social beliefs evolve in response to events -- from policy changes to scientific breakthroughs -- remains a fundamental challenge in social science. Given LLMs' commonsense knowledge and social intelligence, we ask: Can LLMs model the dynamics of social beliefs following social events? In this work, we introduce the concept of the Social World Model (SWM), a general framework designed to capture how social beliefs evolve in response to major events. SWM learns state-transition functions for social beliefs by mining temporal patterns in social data and optimizing the evidence lower bound, without the need for explicit human annotations linking events to belief shifts, or for expensive census data. To evaluate SWM, we introduce a benchmark, SWM-bench, derived from real-world prediction markets, specifically Kalshi and Polymarket. SWM-bench includes over 12k data points for social belief prediction tasks spanning diverse domains such as politics, finance, and cryptocurrency. Our experimental results show that SWM significantly outperforms time-series foundation models, achieving state-of-the-art results on Kalshi data and demonstrating competitive performance on Polymarket data, while offering interpretable insights into the underlying mechanisms of social belief dynamics.
Abstract:Understanding and predicting how social beliefs evolve in response to events -- from policy changes to scientific breakthroughs -- remains a fundamental challenge in social science. Given LLMs' commonsense knowledge and social intelligence, we ask: Can LLMs model the dynamics of social beliefs following social events? In this work, we introduce the concept of the Social World Model (SWM), a general framework designed to capture how social beliefs evolve in response to major events. SWM learns state-transition functions for social beliefs by mining temporal patterns in social data and optimizing the evidence lower bound, without the need for explicit human annotations linking events to belief shifts, or for expensive census data. To evaluate SWM, we introduce a benchmark, SWM-bench, derived from real-world prediction markets, specifically Kalshi and Polymarket. SWM-bench includes over 12k data points for social belief prediction tasks spanning diverse domains such as politics, finance, and cryptocurrency. Our experimental results show that SWM significantly outperforms time-series foundation models, achieving state-of-the-art results on Kalshi data and demonstrating competitive performance on Polymarket data, while offering interpretable insights into the underlying mechanisms of social belief dynamics.
9 pages. ICML 2026
Social and Information Networks (cs.SI); Computation and Language (cs.CL)
arXiv:2606.11482 [cs.SI]
(or arXiv:2606.11482v1 [cs.SI] for this version)
https://doi.org/10.48550/arXiv.2606.11482 Focus to learn more arXiv-issued DOI via DataCite (pending registration)
arXivLabs is a framework that allows collaborators to develop and share new arXiv features directly on our website.
Both individuals and organizations that work with arXivLabs have embraced and accepted our values of openness, community, excellence, and user data privacy. arXiv is committed to these values and only works with partners that adhere to them.
Have an idea for a project that will add value for arXiv's community? Learn more about arXivLabs.
Full text under the publisher's open licence · 288 words
Related coverage
Modular Monolingual Adaptation using Pretrained Language Models
arXiv · Jun 8, 2026
5 Books That Will Deepen Your Understanding of Large Language Models
KDnuggets · Jul 31, 2026
5 Must-Read Resources for Mastering Small Language Models
KDnuggets · Jul 29, 2026
Linux Foundation Launches BUILD Foundation for Large-Scale, Multi-Language Build and Test Systems
HPCwire · Oct 14, 2026