Research Notes

Essays and benchmarks on multi-agent alignment and simulated world challenges.

MAY 2026 • ESSAY #04

Multi-Agent Consensus in High-Stakes Financial Simulations

How synthetic environments enable testing complex AI behavior boundaries without risking real-world capital.

APR 2026 • ESSAY #03

Evaluating Emergent Behaviors in Simulated White-Collar Work

Benchmarking collaborative agent fleets across multi-step research, code generation, and verification workflows.