Definition
Plain language
An agentic scientific-computing framework that grows memory of methods as it solves problems.
As stated in the literature
GRAFT-ATHENA, an agentic scientific computing framework using a geometric factored action tree, problem fingerprinting, and similarity-weighted memory for warm-starting new problems.
Also called: GRAFT-ATHENA
Why it matters: Reusing what worked on related problems is how human scientists actually operate, and building it into an agent gets compounding returns over many tasks.
For example, when GRAFT encounters a new differential equation, it looks up similar past problems in its memory and warm-starts with the integration methods that worked there.
Heard on the show
“jl documentation directly without the GRAFT substrate.”Episode 042 — An Agentic Scientific Computing System That Actually Remembers What It Learns