Tag
Strabo is a research contribution that models Google's Universal Commerce Protocol (UCP) as a declarative Langshaw protocol and implements agents using the Peach programming model, demonstrating interoperability between formally-specified agents and Google's UCP agents for agentic AI e-commerce interactions.
This paper identifies vulnerabilities in the AIVAT variance reduction technique when the heuristic value function is not fixed prior to evaluation, and shows how to propagate heuristic uncertainty to further reduce variance, achieving a 43% reduction in the number of samples needed for statistical conclusions.
OpenAI researchers propose a general framework for learning representations of agent policies in multiagent systems using minimal interaction data, casting the problem as representation learning with applications to competitive control and cooperative communication environments.