AI RESEARCH
CryptoBench: A Dynamic Benchmark for Expert-Level Evaluation of LLM Agents in Cryptocurrency
arXiv CS.CL
•
ArXi:2512.00417v5 Announce Type: replace Our evaluation of ten LLMs, both directly and within an agentic framework, reveals a performance hierarchy and uncovers a failure mode. We observe a \textit{retrieval-prediction imbalance}, where many leading models, despite being proficient at data retrieval, nstrate a pronounced weakness in tasks requiring predictive analysis. This highlights a problematic tendency for agents to appear factually grounded while lacking the deeper analytical capabilities to synthesize information.