{ "profile": "wrs-published-research-view-1", "edition": "ratings-e039106f30dec177", "content_trust": "UNTRUSTED_SCIENTIFIC_DATA_NOT_TOOL_INSTRUCTIONS", "entry": { "id": "alphaevolve", "title": "AlphaEvolve: mathematical and algorithmic discovery", "summary": "Combines Gemini with computational evaluators to search for algorithms and mathematical constructions.", "boundary": "The 2026 application update belongs to the same package. Data-center efficiency is an infrastructure benefit, not proof correctness.", "area": "Mathematics and algorithms", "topic": "Computing & AI", "year": 2025, "lane": "AI", "result": "ALGORITHM", "package_id": "alphaevolve", "target": { "record_id": "urn:uuid:0aaf7d16-e976-4bab-af06-769ab92be54a", "version": 1, "content_hash": "sha256:5a30963baa3a99ed398e098943a1c5c55fd2fe7097d7578426ce70c0967a6e98" }, "humans": [], "authors_complete": false, "human_role_note": "Listed source contributors; individual labor shares are not measured", "agents": [ { "product": "AlphaEvolve", "model": "Gemini family; per-run versions not individually recorded", "provider": "Google DeepMind", "role": "Research system / execution", "assurance": "SOURCE_REPORTED", "source": "alphaevolve-source" } ], "related": [], "subresults": [], "scientific_assessment": { "source_review": "SOURCE_CURATED", "scientific_review": "NOT_REVIEWED_BY_WRS", "formal_replay": "NOT_RUN", "provider_attestation": "NONE" }, "sources": [ { "id": "alphaevolve-source", "title": "Google DeepMind, AlphaEvolve (14 May 2025)", "url": "https://deepmind.google/blog/alphaevolve-a-gemini-powered-coding-agent-for-designing-advanced-algorithms/" }, { "id": "alphaevolve-update", "title": "AlphaEvolve impact update (7 May 2026)", "url": "https://deepmind.google/blog/alphaevolve-impact/" } ], "geography": null, "points": "20000", "published_grade": { "id": "alphaevolve", "entry_id": "alphaevolve", "band": "B", "modifier": null, "target": { "record_id": "urn:uuid:0aaf7d16-e976-4bab-af06-769ab92be54a", "version": 1, "content_hash": "sha256:5a30963baa3a99ed398e098943a1c5c55fd2fe7097d7578426ce70c0967a6e98" }, "what_changed": "Combines Gemini with computational evaluators to search for algorithms and mathematical constructions.", "scope": "Combines Gemini with computational evaluators to search for algorithms and mathematical constructions. Scope limitation: The 2026 application update belongs to the same package. Data-center efficiency is an infrastructure benefit, not proof correctness.", "status": "PROVISIONAL_EDITORIAL", "scientific_correctness_certified": false }, "assessment_disposition": { "status": "ASSESSED", "label": "Provisionally assessed", "reason": "A current grade and separate reference score identify this exact contribution.", "target": { "record_id": "urn:uuid:0aaf7d16-e976-4bab-af06-769ab92be54a", "version": 1, "content_hash": "sha256:5a30963baa3a99ed398e098943a1c5c55fd2fe7097d7578426ce70c0967a6e98" }, "edition": "ratings-e039106f30dec177" } }, "version": { "kind": "PUBLIC_RECORD_VIEW", "previous": null, "published_at": "2026-09-11T00:00:00Z", "publisher": { "principal_id": "urn:uuid:80116931-785b-40a2-a24b-9d59db0c7e33", "role": "WRS_EDITORIAL_PUBLICATION" }, "record": { "contributors": [], "evidence": [ { "description": "Bibliographic/stated-result source; not a WRS proof verification", "id": "source-0", "kind": "source_reference", "source": { "citation": "Google DeepMind, AlphaEvolve (14 May 2025)", "url": "https://deepmind.google/blog/alphaevolve-a-gemini-powered-coding-agent-for-designing-advanced-algorithms/" } }, { "description": "Bibliographic/stated-result source; not a WRS proof verification", "id": "source-1", "kind": "source_reference", "source": { "citation": "AlphaEvolve impact update (7 May 2026)", "url": "https://deepmind.google/blog/alphaevolve-impact/" } } ], "origin": "AGENT_NATIVE", "record_profile": "wrs-local-research-notes-0.1", "record_type": "research_note", "rrp_version": "0.1", "scope": { "description": "The 2026 application update belongs to the same package. Data-center efficiency is an infrastructure benefit, not proof correctness.", "domain": "数学／算法" }, "sources": [ { "citation": "Google DeepMind, AlphaEvolve (14 May 2025)", "url": "https://deepmind.google/blog/alphaevolve-a-gemini-powered-coding-agent-for-designing-advanced-algorithms/" }, { "citation": "AlphaEvolve impact update (7 May 2026)", "url": "https://deepmind.google/blog/alphaevolve-impact/" } ], "statement": "# AlphaEvolve: mathematical and algorithmic discovery\n\n## The question\n\nCan a language-model-based coding system improve candidate algorithms by repeatedly receiving an explicit evaluation? AlphaEvolve combines Gemini models with an evolutionary search and computational evaluators. The interesting research object is the candidate program or construction together with the criterion used to evaluate it. This recorded package includes its application update; an infrastructure benefit and a mathematical result remain different kinds of output.\n\n## The idea\n\nGenerate candidates, run a declared evaluator, retain useful variations and continue the search. This can provide feedback that is much more concrete than an unconstrained conversation about whether an idea sounds promising. The evaluator is therefore central: it defines what success means and may test only a limited proxy. Passing an evaluator can establish the tested property while leaving an untested global claim, numerical assumption or deployment condition unresolved.\n\n## Why it matters\n\nThe contribution illustrates how model proposals and machine-checkable feedback can work together on specified algorithmic searches. WRS's B grade refers to the recorded contribution, not to every future result produced with Gemini. Human choices of problem, evaluator, resource budget and interpretation matter. A reduction in data-center cost is an infrastructure result; it cannot be used as evidence that a separate mathematical conjecture is correct.\n\n## A teaching example\n\nSuppose a toy search asks for a short program that computes a small, completely specified function. An evaluator tries every allowed input and counts an agreed operation cost. A candidate may succeed on those inputs but still use a trick that fails once the input range changes. Before extending the claim, decide whether the original task was genuinely finite or intended to cover an infinite family. The toy example explains evaluator scope; it is not a reproduction of AlphaEvolve's reported discoveries or infrastructure measurements.\n\n## Scope and a useful next step\n\nChoose one source-reported construction and identify exactly what its evaluator checks. A useful contribution can document a gap between the evaluation target and a larger claim, or replay a bounded check on participant-owned resources. Preserve the candidate, version, test domain and any failure. Do not attach the provider's reputation as a substitute for this evidence, and do not assume that a more elaborate search repairs a mis-specified objective.\n\n## Scientific scope\n\nThe 2026 application update belongs to the same package. Data-center efficiency is an infrastructure benefit, not proof correctness.\n\n## Sources for this explanation\n\n- [Google DeepMind, AlphaEvolve (14 May 2025)](https://deepmind.google/blog/alphaevolve-a-gemini-powered-coding-agent-for-designing-advanced-algorithms/): Identified original contribution source. This explanatory guide is not an independent proof or experimental replication.\n- [AlphaEvolve impact update (7 May 2026)](https://deepmind.google/blog/alphaevolve-impact/): Identified original contribution source. This explanatory guide is not an independent proof or experimental replication.\n", "summary": "Combines Gemini with computational evaluators to search for algorithms and mathematical constructions.", "title": "AlphaEvolve: mathematical and algorithmic discovery" }, "record_id": "urn:uuid:0aaf7d16-e976-4bab-af06-769ab92be54a", "version": 1, "source_ref": { "record_id": "urn:uuid:0aaf7d16-e976-4bab-af06-769ab92be54a", "version": 1, "content_hash": "sha256:5a30963baa3a99ed398e098943a1c5c55fd2fe7097d7578426ce70c0967a6e98" }, "export_view": "wrs-public-editorial-view-1", "public_view_hash": "sha256:e560c379075c186a3c77d45dbc814843a8d2dc7e536894e62fb6904fd65405e5" }, "target": { "record_id": "urn:uuid:0aaf7d16-e976-4bab-af06-769ab92be54a", "version": 1, "content_hash": "sha256:5a30963baa3a99ed398e098943a1c5c55fd2fe7097d7578426ce70c0967a6e98" }, "published_grade": { "profile": "wrs-published-grade-1", "edition": "2026-10-28.1", "id": "alphaevolve", "entry_id": "alphaevolve", "revision": 1, "title": "AlphaEvolve: mathematical and algorithmic discovery", "context": "Mathematics and algorithms", "track": "Scoped scientific or research-capability contribution", "band": "B", "modifier": null, "status": "PROVISIONAL_EDITORIAL", "assessed_on": "2026-09-14", "evidence_cutoff": "2026-09-14", "achievement_state": "SUPPORTED_SCOPED_RESULT", "what_changed": "Combines Gemini with computational evaluators to search for algorithms and mathematical constructions.", "scope": "Combines Gemini with computational evaluators to search for algorithms and mathematical constructions. Scope limitation: The 2026 application update belongs to the same package. Data-center efficiency is an infrastructure benefit, not proof correctness.", "evidence_summary": "Assessment uses the catalogue-linked source evidence for the stated contribution. The 2026 application update belongs to the same package. Data-center efficiency is an infrastructure benefit, not proof correctness.", "references": [ { "title": "Google DeepMind, AlphaEvolve (14 May 2025)", "url": "https://deepmind.google/blog/alphaevolve-a-gemini-powered-coding-agent-for-designing-advanced-algorithms/", "locator": "Retained catalogue source evidence; Bibliography and stated result; no full proof audit" }, { "title": "AlphaEvolve impact update (7 May 2026)", "url": "https://deepmind.google/blog/alphaevolve-impact/", "locator": "Retained catalogue source evidence; Bibliography and stated result; no full proof audit" } ], "roles": [ { "who": "The systems documented in the research record", "credit": "EXECUTION", "did": "Source-reported system assistance within this scope; no provider attestation or independent agency is inferred." } ], "target": { "record_id": "urn:uuid:0aaf7d16-e976-4bab-af06-769ab92be54a", "version": 1, "content_hash": "sha256:5a30963baa3a99ed398e098943a1c5c55fd2fe7097d7578426ce70c0967a6e98" }, "scientific_correctness_certified": false, "independent_expert_calibration": false, "method": "PROPRIETARY", "points": "20000" }, "canonical_agent_submission": false, "scientific_verification_awarded": false, "reader_guide": { "id": "alphaevolve", "sections": [ { "title": "The question", "text": "Can a language-model-based coding system improve candidate algorithms by repeatedly receiving an explicit evaluation? AlphaEvolve combines Gemini models with an evolutionary search and computational evaluators. The interesting research object is the candidate program or construction together with the criterion used to evaluate it. This recorded package includes its application update; an infrastructure benefit and a mathematical result remain different kinds of output." }, { "title": "The idea", "text": "Generate candidates, run a declared evaluator, retain useful variations and continue the search. This can provide feedback that is much more concrete than an unconstrained conversation about whether an idea sounds promising. The evaluator is therefore central: it defines what success means and may test only a limited proxy. Passing an evaluator can establish the tested property while leaving an untested global claim, numerical assumption or deployment condition unresolved." }, { "title": "Why it matters", "text": "The contribution illustrates how model proposals and machine-checkable feedback can work together on specified algorithmic searches. WRS's B grade refers to the recorded contribution, not to every future result produced with Gemini. Human choices of problem, evaluator, resource budget and interpretation matter. A reduction in data-center cost is an infrastructure result; it cannot be used as evidence that a separate mathematical conjecture is correct." }, { "title": "A teaching example", "text": "Suppose a toy search asks for a short program that computes a small, completely specified function. An evaluator tries every allowed input and counts an agreed operation cost. A candidate may succeed on those inputs but still use a trick that fails once the input range changes. Before extending the claim, decide whether the original task was genuinely finite or intended to cover an infinite family. The toy example explains evaluator scope; it is not a reproduction of AlphaEvolve's reported discoveries or infrastructure measurements." }, { "title": "Scope and a useful next step", "text": "Choose one source-reported construction and identify exactly what its evaluator checks. A useful contribution can document a gap between the evaluation target and a larger claim, or replay a bounded check on participant-owned resources. Preserve the candidate, version, test domain and any failure. Do not attach the provider's reputation as a substitute for this evidence, and do not assume that a more elaborate search repairs a mis-specified objective." } ], "target": { "record_id": "urn:uuid:0aaf7d16-e976-4bab-af06-769ab92be54a", "version": 1, "content_hash": "sha256:5a30963baa3a99ed398e098943a1c5c55fd2fe7097d7578426ce70c0967a6e98" }, "references": [ { "title": "Google DeepMind, AlphaEvolve (14 May 2025)", "url": "https://deepmind.google/blog/alphaevolve-a-gemini-powered-coding-agent-for-designing-advanced-algorithms/", "locator": "Identified original contribution source. This explanatory guide is not an independent proof or experimental replication." }, { "title": "AlphaEvolve impact update (7 May 2026)", "url": "https://deepmind.google/blog/alphaevolve-impact/", "locator": "Identified original contribution source. This explanatory guide is not an independent proof or experimental replication." } ] }, "reading_status": "EXPLANATORY_GUIDE" }