Record · analyzed
Darwin Gödel Machine: Open-Ended Evolution of Self-Improving Agents
Keeps an archive of agent versions; each new version is an agent rewriting its own code, kept if it scores well: SWE-bench 20% → 50%.
1
Listed
Added to the list. How entries were checked when they were listed is in the journal.
2
Collected
| File | Source | SHA-256 |
|---|---|---|
| paper PDF, 2505.22954v32505.22954.pdf · 3.8 MB | arxiv.org/pdf/2505.22954v3 | 13ff4abe0c7ad4a7dd3b4876d19a8bf940e39e70dabbf06065aa774a6c3457de |
To check a file, download it from its source and compare its SHA-256 with the one above. Collected copies are kept in the editor's RSI system; manifest: design-life/literature/rsi-design/sources.json.
3
Analyzed
Quoted verbatim from design-life/literature/rsi-workflow-improvement/comparison.md (2026-10-07); design-life/research/rsi-design/evidence-review-2026-10-28.md (2026-10-28). Parts about applying the paper to the editor's private work are left out.
From “Evidence for improving reusable agent workflows” (2026-10-07)
Inspected edition: Darwin Gödel Machine, v3, 12 March 2026; PDF pp. 3–7
Mechanism and reported evidence: Edits an agent around a frozen model, tests variants and retains an archive. After 80 iterations, reported SWE-bench Verified success rises from 20% to 50% under its staged subset protocol; full Polyglot pass@1 rises from 14.2% to 30.7%.
From “Evidence for effective RSI implementation” (2026-10-28)
The evidence also leaves other productive changes available. DGM changes agent tools and software under fixed task criteria. AIDE² changes search and context policies. ACE changes reusable knowledge contents. AFlow changes prompts and connections. These interventions overlap, but they should not all be relabeled as improvements to the semantic task schema. DGM
4
Replicated
Not done yet. Not attempted yet.
5
Audited
Not done yet. Not attempted yet.
The editor's view
None recorded yet.