World Wires · story 28549 · corroborated · 1 source(s)

LiveMACE: Process-Aware Evaluation of LLM Agents in Markets

LiveMACEBench evaluates persistent large language model agent capabilities using live financial markets as an evolving testbed.

Open in the desk

Coverage

• LiveMACE: Process-Aware Evaluation of LLM Agents in Markets arXiv: new preprints (security, AI, physics) · tier C · clear net · 2026-10-07 11:33 UTC

What this site indexes