2 papers
cs.AI2026
FORGE: Self-Evolving Agent Memory With No Weight Updates via Population Broadcast
Igor Bogdanov, Chung-Horng Lung, Thomas Kunz +3
Can LLM agents improve decision-making through self-generated memory without gradient updates? We propose FORGE (Failure-Optimized Reflective Graduation and Evolution), a staged, p…
cs.AI2026
Context, Reasoning, and Hierarchy: A Cost-Performance Study of Compound LLM Agent Design in an Adversarial POMDP
Igor Bogdanov, Chung-Horng Lung, Thomas Kunz +3
Deploying compound LLM agents in adversarial, partially observable sequential environments requires navigating several design dimensions: (1) what the agent sees, (2) how it reason…