AI RESEARCH

OracleProto: A Reproducible Framework for Benchmarking LLM Native Forecasting via Knowledge Cutoff and Temporal Masking

arXiv CS.AI

ArXi:2605.03762v1 Announce Type: new Large language models are moving from static text generators toward real-world decision- systems, where forecasting is a composite capability that links information gathering, evidence integration, situational judgment, and action-oriented decision making.