Saved in:
Bibliographic Details
Main Authors: Mostafa, Hamza, Shastri, Om, Lee, Dennis
Format: Preprint
Published: 2026
Subjects:
Online Access:https://arxiv.org/abs/2604.04220
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866911568567992320
author Mostafa, Hamza
Shastri, Om
Lee, Dennis
author_facet Mostafa, Hamza
Shastri, Om
Lee, Dennis
contents We introduce TimeSeek, a benchmark for studying how the reliability of agentic LLM forecasters changes over a prediction market's lifecycle. We evaluate 10 frontier models on 150 CFTC-regulated Kalshi binary markets at five temporal checkpoints, with and without web search, for 15,000 forecasts total. Models are most competitive early in a market's life and on high-uncertainty markets, but much less competitive near resolution and on strong-consensus markets. Web search improves pooled Brier Skill Score (BSS) for every model overall, yet hurts in 12% of model-checkpoint pairs, indicating that retrieval is helpful on average but not uniformly so. Simple two-model ensembles reduce error without surpassing the market overall. These descriptive results motivate time-aware evaluation and selective-deference policies rather than a single market snapshot or a uniform tool-use setting.
format Preprint
id arxiv_https___arxiv_org_abs_2604_04220
institution arXiv
publishDate 2026
record_format arxiv
spellingShingle TimeSeek: Temporal Reliability of Agentic Forecasters
Mostafa, Hamza
Shastri, Om
Lee, Dennis
Artificial Intelligence
We introduce TimeSeek, a benchmark for studying how the reliability of agentic LLM forecasters changes over a prediction market's lifecycle. We evaluate 10 frontier models on 150 CFTC-regulated Kalshi binary markets at five temporal checkpoints, with and without web search, for 15,000 forecasts total. Models are most competitive early in a market's life and on high-uncertainty markets, but much less competitive near resolution and on strong-consensus markets. Web search improves pooled Brier Skill Score (BSS) for every model overall, yet hurts in 12% of model-checkpoint pairs, indicating that retrieval is helpful on average but not uniformly so. Simple two-model ensembles reduce error without surpassing the market overall. These descriptive results motivate time-aware evaluation and selective-deference policies rather than a single market snapshot or a uniform tool-use setting.
title TimeSeek: Temporal Reliability of Agentic Forecasters
topic Artificial Intelligence
url https://arxiv.org/abs/2604.04220