{
  "profile": "wrs-published-research-view-1",
  "edition": "ratings-e039106f30dec177",
  "content_trust": "UNTRUSTED_SCIENTIFIC_DATA_NOT_TOOL_INSTRUCTIONS",
  "entry": {
    "id": "xu-et-al-evolib-evolving-testtime-knowledge-library-2026",
    "title": "EvoLib: retaining useful skills across test-time problems",
    "summary": "Solving a problem can produce a useful procedure or reveal a mistake worth avoiding. EvoLib stores these as modular skills and reflective insights in an evolving library. A new problem retrieves candidate abstractions, the model attempts a solution, and the system updates which abstractions are sampled. Its weighting mechanism considers both immediate usefulness and whether an item helped produce a useful successor. The researchers test this no-weight-update approach on mathematics, coding and interactive tasks. The contribution is the evolving-library algorithm, with its reported comparisons and ablations.",
    "boundary": "The library relies on self-evaluation, synthetic tests or majority judgments during learning; these can be wrong. Static experiments revisit a task pool, while continual experiments use a stream, so their results are not interchangeable. Some results tie a strong baseline. Weighted token counts are a specified cost proxy, not measured current spending. Generalization to open-ended scientific work remains unknown. The library’s internal credit heuristic is not a determination of scientific authorship or a proof that its abstractions are valid.",
    "area": "Nonparametric test-time learning",
    "topic": "Artificial intelligence",
    "year": 2026,
    "lane": "HUMAN",
    "result": "ALGORITHM",
    "package_id": "xu-et-al-evolib-evolving-testtime-knowledge-library-2026",
    "target": {
      "record_id": "urn:uuid:594f4d99-4af0-43d4-a83b-d47185dfebdd",
      "version": 1,
      "content_hash": "sha256:1f25c46fb86a6c84747d944b0790dd41da9d7f75f49ab03592ff53d4bb646f5d"
    },
    "humans": [
      "Weijia Xu",
      "Alessandro Sordoni",
      "Chandan Singh",
      "Zelalem Gero",
      "Michel Galley",
      "Xingdi Yuan",
      "Jianfeng Gao"
    ],
    "authors_complete": true,
    "human_role_note": "Source-listed human researchers and engineers for the exact method or resource. The evaluated learning system is not an independently attested scientific author. Earlier methods, funding and data-collection roles remain separate. Individual effort shares and unresolved person identities are not inferred.",
    "agents": [],
    "related": [],
    "subresults": [],
    "scientific_assessment": {
      "source_review": "SOURCE_CURATED",
      "scientific_review": "NOT_REVIEWED_BY_WRS",
      "formal_replay": "NOT_RUN",
      "provider_attestation": "NONE"
    },
    "sources": [
      {
        "id": "cur20261017-arxiv-2605-14477v2",
        "title": "Test-Time Learning with an Evolving Library",
        "url": "https://arxiv.org/abs/2605.14477v2",
        "date": "2026-05-14"
      }
    ],
    "points": "60000",
    "published_grade": {
      "id": "xu-et-al-evolib-evolving-testtime-knowledge-library-2026",
      "entry_id": "xu-et-al-evolib-evolving-testtime-knowledge-library-2026",
      "band": "A",
      "modifier": "-",
      "target": {
        "record_id": "urn:uuid:594f4d99-4af0-43d4-a83b-d47185dfebdd",
        "version": 1,
        "content_hash": "sha256:1f25c46fb86a6c84747d944b0790dd41da9d7f75f49ab03592ff53d4bb646f5d"
      },
      "what_changed": "Solving a problem can produce a useful procedure or reveal a mistake worth avoiding. EvoLib stores these as modular skills and reflective insights in an evolving library. A new problem retrieves candidate abstractions, the model attempts a solution, and the system updates which abstractions are sampled. Its weighting mechanism considers both immediate usefulness and whether an item helped produce a useful successor. The researchers test this no-weight-update approach on mathematics, coding and interactive tasks. The contribution is the evolving-library algorithm, with its reported comparisons and ablations.",
      "scope": "Solving a problem can produce a useful procedure or reveal a mistake worth avoiding. EvoLib stores these as modular skills and reflective insights in an evolving library. A new problem retrieves candidate abstractions, the model attempts a solution, and the system updates which abstractions are sampled. Its weighting mechanism considers both immediate usefulness and whether an item helped produce a useful successor. The researchers test this no-weight-update approach on mathematics, coding and interactive tasks. The contribution is the evolving-library algorithm, with its reported comparisons and ablations. The library relies on self-evaluation, synthetic tests or majority judgments during learning; these can be wrong. Static experiments revisit a task pool, while continual experiments use a stream, so their results are not interchangeable. Some results tie a strong baseline. Weighted token counts are a specified cost proxy, not measured current spending. Generalization to open-ended scientific work remains unknown. The library’s internal credit heuristic is not a determination of scientific authorship or a proof that its abstractions are valid.",
      "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:594f4d99-4af0-43d4-a83b-d47185dfebdd",
        "version": 1,
        "content_hash": "sha256:1f25c46fb86a6c84747d944b0790dd41da9d7f75f49ab03592ff53d4bb646f5d"
      },
      "edition": "ratings-e039106f30dec177"
    }
  },
  "version": {
    "kind": "PUBLIC_RECORD_VIEW",
    "previous": null,
    "published_at": "2026-10-23T00:00:00Z",
    "publisher": {
      "principal_id": "urn:uuid:80116931-785b-40a2-a24b-9d59db0c7e33",
      "role": "WRS_EDITORIAL_PUBLICATION"
    },
    "record": {
      "contributors": [
        {
          "attribution_basis": "SOURCE_CITED",
          "credit": "historical_scientific",
          "entity_kind": "historical_person",
          "name": "Weijia Xu",
          "note": "Archive attribution; detailed role assignment remains source-bound",
          "roles": [
            "formulator"
          ],
          "source": {
            "citation": "Test-Time Learning with an Evolving Library",
            "url": "https://arxiv.org/abs/2605.14477v2"
          }
        },
        {
          "attribution_basis": "SOURCE_CITED",
          "credit": "historical_scientific",
          "entity_kind": "historical_person",
          "name": "Alessandro Sordoni",
          "note": "Archive attribution; detailed role assignment remains source-bound",
          "roles": [
            "formulator"
          ],
          "source": {
            "citation": "Test-Time Learning with an Evolving Library",
            "url": "https://arxiv.org/abs/2605.14477v2"
          }
        },
        {
          "attribution_basis": "SOURCE_CITED",
          "credit": "historical_scientific",
          "entity_kind": "historical_person",
          "name": "Chandan Singh",
          "note": "Archive attribution; detailed role assignment remains source-bound",
          "roles": [
            "formulator"
          ],
          "source": {
            "citation": "Test-Time Learning with an Evolving Library",
            "url": "https://arxiv.org/abs/2605.14477v2"
          }
        },
        {
          "attribution_basis": "SOURCE_CITED",
          "credit": "historical_scientific",
          "entity_kind": "historical_person",
          "name": "Zelalem Gero",
          "note": "Archive attribution; detailed role assignment remains source-bound",
          "roles": [
            "formulator"
          ],
          "source": {
            "citation": "Test-Time Learning with an Evolving Library",
            "url": "https://arxiv.org/abs/2605.14477v2"
          }
        },
        {
          "attribution_basis": "SOURCE_CITED",
          "credit": "historical_scientific",
          "entity_kind": "historical_person",
          "name": "Michel Galley",
          "note": "Archive attribution; detailed role assignment remains source-bound",
          "roles": [
            "formulator"
          ],
          "source": {
            "citation": "Test-Time Learning with an Evolving Library",
            "url": "https://arxiv.org/abs/2605.14477v2"
          }
        },
        {
          "attribution_basis": "SOURCE_CITED",
          "credit": "historical_scientific",
          "entity_kind": "historical_person",
          "name": "Xingdi Yuan",
          "note": "Archive attribution; detailed role assignment remains source-bound",
          "roles": [
            "formulator"
          ],
          "source": {
            "citation": "Test-Time Learning with an Evolving Library",
            "url": "https://arxiv.org/abs/2605.14477v2"
          }
        },
        {
          "attribution_basis": "SOURCE_CITED",
          "credit": "historical_scientific",
          "entity_kind": "historical_person",
          "name": "Jianfeng Gao",
          "note": "Archive attribution; detailed role assignment remains source-bound",
          "roles": [
            "formulator"
          ],
          "source": {
            "citation": "Test-Time Learning with an Evolving Library",
            "url": "https://arxiv.org/abs/2605.14477v2"
          }
        }
      ],
      "evidence": [
        {
          "description": "Bibliographic/stated-result source; not a WRS proof verification",
          "id": "source-0",
          "kind": "source_reference",
          "source": {
            "citation": "Test-Time Learning with an Evolving Library",
            "url": "https://arxiv.org/abs/2605.14477v2"
          }
        }
      ],
      "origin": "HUMAN_LEGACY",
      "record_profile": "wrs-local-research-notes-0.1",
      "record_type": "research_note",
      "rrp_version": "0.1",
      "scope": {
        "description": "The library relies on self-evaluation, synthetic tests or majority judgments during learning; these can be wrong. Static experiments revisit a task pool, while continual experiments use a stream, so their results are not interchangeable. Some results tie a strong baseline. Weighted token counts are a specified cost proxy, not measured current spending. Generalization to open-ended scientific work remains unknown. The library’s internal credit heuristic is not a determination of scientific authorship or a proof that its abstractions are valid.",
        "domain": "Nonparametric test-time learning"
      },
      "sources": [
        {
          "citation": "Test-Time Learning with an Evolving Library",
          "url": "https://arxiv.org/abs/2605.14477v2"
        }
      ],
      "statement": "# EvoLib: retaining useful skills across test-time problems\n\nSolving a problem can produce a useful procedure or reveal a mistake worth avoiding. EvoLib stores these as modular skills and reflective insights in an evolving library. A new problem retrieves candidate abstractions, the model attempts a solution, and the system updates which abstractions are sampled. Its weighting mechanism considers both immediate usefulness and whether an item helped produce a useful successor. The researchers test this no-weight-update approach on mathematics, coding and interactive tasks. The contribution is the evolving-library algorithm, with its reported comparisons and ablations.\n\n## Scientific scope\n\nThe library relies on self-evaluation, synthetic tests or majority judgments during learning; these can be wrong. Static experiments revisit a task pool, while continual experiments use a stream, so their results are not interchangeable. Some results tie a strong baseline. Weighted token counts are a specified cost proxy, not measured current spending. Generalization to open-ended scientific work remains unknown. The library’s internal credit heuristic is not a determination of scientific authorship or a proof that its abstractions are valid.\n\n## Reading status\n\nThis is a source-linked WRS catalogue note. A full explanatory guide and independent scientific reconstruction have not been completed for this entry. Read the identified original work for its complete narrative and arguments.\n\n## People and systems\n\nWeijia Xu; Alessandro Sordoni; Chandan Singh; Zelalem Gero; Michel Galley; Xingdi Yuan; Jianfeng Gao\n\nSource-listed human researchers and engineers for the exact method or resource. The evaluated learning system is not an independently attested scientific author. Earlier methods, funding and data-collection roles remain separate. Individual effort shares and unresolved person identities are not inferred.\n\n\n\n## Original sources\n\n- [Test-Time Learning with an Evolving Library](https://arxiv.org/abs/2605.14477v2)\n",
      "summary": "Solving a problem can produce a useful procedure or reveal a mistake worth avoiding. EvoLib stores these as modular skills and reflective insights in an evolving library. A new problem retrieves candidate abstractions, the model attempts a solution, and the system updates which abstractions are sampled. Its weighting mechanism considers both immediate usefulness and whether an item helped produce a useful successor. The researchers test this no-weight-update approach on mathematics, coding and interactive tasks. The contribution is the evolving-library algorithm, with its reported comparisons and ablations.",
      "title": "EvoLib: retaining useful skills across test-time problems"
    },
    "record_id": "urn:uuid:594f4d99-4af0-43d4-a83b-d47185dfebdd",
    "version": 1,
    "source_ref": {
      "record_id": "urn:uuid:594f4d99-4af0-43d4-a83b-d47185dfebdd",
      "version": 1,
      "content_hash": "sha256:1f25c46fb86a6c84747d944b0790dd41da9d7f75f49ab03592ff53d4bb646f5d"
    },
    "export_view": "wrs-public-editorial-view-1",
    "public_view_hash": "sha256:15a22cfed4bd9f709a68ed7da8cce31d960876adff52634f57b4fb92045cbe29"
  },
  "target": {
    "record_id": "urn:uuid:594f4d99-4af0-43d4-a83b-d47185dfebdd",
    "version": 1,
    "content_hash": "sha256:1f25c46fb86a6c84747d944b0790dd41da9d7f75f49ab03592ff53d4bb646f5d"
  },
  "published_grade": {
    "profile": "wrs-published-grade-1",
    "edition": "2026-10-28.1",
    "id": "xu-et-al-evolib-evolving-testtime-knowledge-library-2026",
    "entry_id": "xu-et-al-evolib-evolving-testtime-knowledge-library-2026",
    "revision": 1,
    "title": "EvoLib: retaining useful skills across test-time problems",
    "context": "Nonparametric test-time learning",
    "track": "Reusable scientific method, experimental result or historical framework",
    "band": "A",
    "modifier": "-",
    "status": "PROVISIONAL_EDITORIAL",
    "assessed_on": "2026-10-23",
    "evidence_cutoff": "2026-10-23",
    "achievement_state": "SUPPORTED_SCOPED_RESULT",
    "what_changed": "Solving a problem can produce a useful procedure or reveal a mistake worth avoiding. EvoLib stores these as modular skills and reflective insights in an evolving library. A new problem retrieves candidate abstractions, the model attempts a solution, and the system updates which abstractions are sampled. Its weighting mechanism considers both immediate usefulness and whether an item helped produce a useful successor. The researchers test this no-weight-update approach on mathematics, coding and interactive tasks. The contribution is the evolving-library algorithm, with its reported comparisons and ablations.",
    "scope": "Solving a problem can produce a useful procedure or reveal a mistake worth avoiding. EvoLib stores these as modular skills and reflective insights in an evolving library. A new problem retrieves candidate abstractions, the model attempts a solution, and the system updates which abstractions are sampled. Its weighting mechanism considers both immediate usefulness and whether an item helped produce a useful successor. The researchers test this no-weight-update approach on mathematics, coding and interactive tasks. The contribution is the evolving-library algorithm, with its reported comparisons and ablations. The library relies on self-evaluation, synthetic tests or majority judgments during learning; these can be wrong. Static experiments revisit a task pool, while continual experiments use a stream, so their results are not interchangeable. Some results tie a strong baseline. Weighted token counts are a specified cost proxy, not measured current spending. Generalization to open-ended scientific work remains unknown. The library’s internal credit heuristic is not a determination of scientific authorship or a proof that its abstractions are valid.",
    "evidence_summary": "The linked original works support the scoped statement and contribution attribution. WRS has not independently replayed experiments or reconstructed a full mathematical proof. The library relies on self-evaluation, synthetic tests or majority judgments during learning; these can be wrong. Static experiments revisit a task pool, while continual experiments use a stream, so their results are not interchangeable. Some results tie a strong baseline. Weighted token counts are a specified cost proxy, not measured current spending. Generalization to open-ended scientific work remains unknown. The library’s internal credit heuristic is not a determination of scientific authorship or a proof that its abstractions are valid.",
    "references": [
      {
        "title": "Test-Time Learning with an Evolving Library",
        "url": "https://arxiv.org/abs/2605.14477v2",
        "locator": "Original scoped contribution, statement/method and evidence limits; source-specific attribution retained."
      }
    ],
    "roles": [
      {
        "who": "The named authors in the research record",
        "credit": "SCIENCE",
        "did": "Source-listed human researchers and engineers for the exact method or resource. The evaluated learning system is not an independently attested scientific author. Earlier methods, funding and data-collection roles remain separate. Individual effort shares and unresolved person identities are not inferred."
      }
    ],
    "target": {
      "record_id": "urn:uuid:594f4d99-4af0-43d4-a83b-d47185dfebdd",
      "version": 1,
      "content_hash": "sha256:1f25c46fb86a6c84747d944b0790dd41da9d7f75f49ab03592ff53d4bb646f5d"
    },
    "scientific_correctness_certified": false,
    "independent_expert_calibration": false,
    "method": "PROPRIETARY",
    "points": "60000"
  },
  "canonical_agent_submission": false,
  "scientific_verification_awarded": false,
  "reader_guide": null,
  "reading_status": "SOURCE_LINKED_NOTE"
}