Think or Remember: The Exchange Rate of Agentic Inference
Why shorter agent histories can cost more: cache reuse changes the price of reasoning, while forgetting changes what an agent solves.
Read the article →Research
Cheap Reward Hacking Detection
A 14M-parameter encoder matches a frontier LLM judge at detecting reward hacking in agent trajectories — at four orders of magnitude lower cost.
Open publication →Coding Harness Inspection
Analyzing the adoption, market share, and configuration patterns of AI coding harnesses by examining approximately 400,000 public GitHub repositories to serve as a diagnostic thermometer for the evolving ecosystem.
Open publication →Projects
tamarillo/theta-spec
Specification for consistent, reproducible configurations for agents and harnesses.
tamarillo/theta
Reference implementation of the theta specification.
REPO
tamarillo-ai/theta
DOCS
theta.tamarillo.ai