BWhat changed
Maintained a diverse archive of coding-agent programs, selected variants for LLM-guided self-modification and evaluated them on coding benchmarks, allowing improvements to the agent harness itself.
Scientific scope
Maintained a diverse archive of coding-agent programs, selected variants for LLM-guided self-modification and evaluated them on coding benchmarks, allowing improvements to the agent harness itself. Frozen foundation models, evolutionary search and coding benchmarks are inherited. Despite its name, the system empirically tests changes rather than proving Gödel-machine utility improvement. Benchmarks do not certify safe or unbounded recursive intelligence growth.
Evidence
The linked original sources describe this contribution and its evidence limits. WRS has not independently reconstructed the full scientific claims. Frozen foundation models, evolutionary search and coding benchmarks are inherited. Despite its name, the system empirically tests changes rather than proving Gödel-machine utility improvement. Benchmarks do not certify safe or unbounded recursive intelligence growth.
Contributors
- Source-listed authors and technical contributors · SCIENCE: Source-listed authors or explicitly credited technical contributors for the cited versions. Family entries retain source-specific attribution; a combined list is not one original-paper byline or equal individual work. Names alone do not establish resolved identities.
- Darwin Godel Machine: Open-Ended Evolution of Self-Improving Agents
- Zhang et al.: archive-based empirical self-modification of coding agents — author/institutional project
Provisional WRS assessment, edition 2026-10-16.1. Independent expert calibration is pending. This grade does not certify correctness or assign ownership.
Full scientific record · Browse contribution grades · Request a correction