{"state_type":"pith_open_graph_state","state_version":"1.0","pith_number":"pith:2026:FO3KV7ZQDPOKII3BE4GNKO4AXN","merge_version":"pith-open-graph-merge-v1","event_count":2,"valid_event_count":2,"invalid_event_count":0,"equivocation_count":0,"current":{"canonical_record":{"metadata":{"abstract_canon_sha256":"162d308e35b3b883d7d75b5eafaf151f1767463bf9dd6178a7379df5d3f1e932","cross_cats_sorted":["cs.LG"],"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.AI","submitted_at":"2026-05-13T17:45:16Z","title_canon_sha256":"6870de1fca72884d63a7ddd2babf69eda7daea3ee841b1a863a1f4dc71eca2d6"},"schema_version":"1.0","source":{"id":"2605.13821","kind":"arxiv","version":1}},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2605.13821","created_at":"2026-05-18T02:44:15Z"},{"alias_kind":"arxiv_version","alias_value":"2605.13821v1","created_at":"2026-05-18T02:44:15Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2605.13821","created_at":"2026-05-18T02:44:15Z"},{"alias_kind":"pith_short_12","alias_value":"FO3KV7ZQDPOK","created_at":"2026-05-18T12:33:37Z"},{"alias_kind":"pith_short_16","alias_value":"FO3KV7ZQDPOKII3B","created_at":"2026-05-18T12:33:37Z"},{"alias_kind":"pith_short_8","alias_value":"FO3KV7ZQ","created_at":"2026-05-18T12:33:37Z"}],"graph_snapshots":[{"event_id":"sha256:df42dc4372c4b225db410d8f805bd581a15c8aa33e60df7e826a099bc298078c","target":"graph","created_at":"2026-05-18T02:44:15Z","signer":{"key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signer_id":"pith.science","signer_type":"pith_registry"},"payload":{"graph_snapshot":{"author_claims":{"count":0,"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57","strong_count":0},"builder_version":"pith-number-builder-2026-05-17-v1","claims":{"count":4,"items":[{"attestation":"unclaimed","claim_id":"C1","kind":"strongest_claim","source":"verdict.strongest_claim","status":"machine_extracted","text":"AEvo outperforms five evolution baselines, achieving a 26 relative improvement over the strongest baseline. Across three open-ended optimization tasks, AEvo further outperforms four evolution baselines and achieves state-of-the-art performance under the same iteration budget."},{"attestation":"unclaimed","claim_id":"C2","kind":"weakest_assumption","source":"verdict.weakest_assumption","status":"machine_extracted","text":"That editing the procedure or agent context via the meta-agent will reliably steer long-horizon evolution without introducing new forms of drift or instability, and that the accumulated context provides sufficient signal for effective edits."},{"attestation":"unclaimed","claim_id":"C3","kind":"one_line_summary","source":"verdict.one_line_summary","status":"machine_extracted","text":"AEvo introduces a meta-agent that edits the evolution procedure or agent context based on accumulated state, outperforming baselines by 26% relative improvement on agentic benchmarks and achieving SOTA on open-ended tasks."},{"attestation":"unclaimed","claim_id":"C4","kind":"headline","source":"verdict.pith_extraction.headline","status":"machine_extracted","text":"AEvo improves agentic evolution by having a meta-agent edit the search procedure or context using accumulated evidence as state."}],"snapshot_sha256":"794639d0c951775fa11b062d6c23d227dad836593e45bc4d5b4365da2032ddce"},"formal_canon":{"evidence_count":0,"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"paper":{"abstract_excerpt":"Agentic evolution has emerged as a powerful paradigm for improving programs, workflows, and scientific solutions by iteratively generating candidates, evaluating them, and using feedback to guide future search. However, existing methods are typically instantiated either as fixed hand-designed procedures that are modular but rigid, or as general-purpose agents that flexibly integrate feedback but can drift in long-horizon evolution. Both forms accumulate rich evidence over time, including candidates, feedback, traces, and failures, yet lack a stable interface for organizing this evidence and re","authors_text":"Bang Liu, Caiyin Yang, Chenglin Wu, Jianhao Ruan, Jiayi Zhang, Jinyu Xiang, Maojia Song, Yiran Peng, Yixi Ouyang, Yongfeng Gu, Yuyu Luo, Zhiguang Han, Zhitao Wang","cross_cats":["cs.LG"],"headline":"AEvo improves agentic evolution by having a meta-agent edit the search procedure or context using accumulated evidence as state.","license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.AI","submitted_at":"2026-05-13T17:45:16Z","title":"Harnessing Agentic Evolution"},"references":{"count":48,"internal_anchors":18,"resolved_work":48,"sample":[{"cited_arxiv_id":"2507.19457","doi":"","is_internal_anchor":true,"ref_index":1,"title":"GEPA: Reflective Prompt Evolution Can Outperform Reinforcement Learning","work_id":"40b60d06-dc1c-4799-b75d-ff1eca653049","year":2025},{"cited_arxiv_id":"","doi":"","is_internal_anchor":false,"ref_index":2,"title":"Claude Code, 2025.https://docs.anthropic.com/en/docs/claude-code/ overview","work_id":"19e13c18-d615-46f9-a3d6-85f21e74b428","year":2025},{"cited_arxiv_id":"","doi":"","is_internal_anchor":false,"ref_index":3,"title":"Anthropic’s Original Performance Take-Home","work_id":"d5464cac-e27f-4954-b2f0-d866552f34f9","year":2026},{"cited_arxiv_id":"","doi":"","is_internal_anchor":false,"ref_index":4,"title":"An improved example for an autoconvolution inequality","work_id":"3f97e52d-37f1-43ed-a66e-3cbb14795c9f","year":2026},{"cited_arxiv_id":"2505.11831","doi":"","is_internal_anchor":true,"ref_index":5,"title":"ARC-AGI-2: A New Challenge for Frontier AI Reasoning Systems","work_id":"2957f2ae-a92d-479e-a1ff-b0b522445d0b","year":2025}],"snapshot_sha256":"da8a065f6b9371025d62c663023c6dc618e743ab03443fc69086eb0beb24b5a0"},"source":{"id":"2605.13821","kind":"arxiv","version":1},"verdict":{"created_at":"2026-05-14T17:54:03.627190Z","id":"cdaba0c8-e297-42f3-b1a6-ab10aa1e9580","model_set":{"reader":"grok-4.3"},"one_line_summary":"AEvo introduces a meta-agent that edits the evolution procedure or agent context based on accumulated state, outperforming baselines by 26% relative improvement on agentic benchmarks and achieving SOTA on open-ended tasks.","pipeline_version":"pith-pipeline@v0.9.0","pith_extraction_headline":"AEvo improves agentic evolution by having a meta-agent edit the search procedure or context using accumulated evidence as state.","strongest_claim":"AEvo outperforms five evolution baselines, achieving a 26 relative improvement over the strongest baseline. Across three open-ended optimization tasks, AEvo further outperforms four evolution baselines and achieves state-of-the-art performance under the same iteration budget.","weakest_assumption":"That editing the procedure or agent context via the meta-agent will reliably steer long-horizon evolution without introducing new forms of drift or instability, and that the accumulated context provides sufficient signal for effective edits."}},"verdict_id":"cdaba0c8-e297-42f3-b1a6-ab10aa1e9580"}}],"author_attestations":[],"timestamp_anchors":[],"storage_attestations":[],"citation_signatures":[],"replication_records":[],"corrections":[],"mirror_hints":[],"record_created":{"event_id":"sha256:850d955cf30f7d95940a1e30cf134f5481c401f232d398911d21607ab8002c2b","target":"record","created_at":"2026-05-18T02:44:15Z","signer":{"key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signer_id":"pith.science","signer_type":"pith_registry"},"payload":{"attestation_state":"computed","canonical_record":{"metadata":{"abstract_canon_sha256":"162d308e35b3b883d7d75b5eafaf151f1767463bf9dd6178a7379df5d3f1e932","cross_cats_sorted":["cs.LG"],"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.AI","submitted_at":"2026-05-13T17:45:16Z","title_canon_sha256":"6870de1fca72884d63a7ddd2babf69eda7daea3ee841b1a863a1f4dc71eca2d6"},"schema_version":"1.0","source":{"id":"2605.13821","kind":"arxiv","version":1}},"canonical_sha256":"2bb6aaff301bdca42361270cd53b80bb57c27d9af8c34cb262684f4e6a034e51","receipt":{"algorithm":"ed25519","builder_version":"pith-number-builder-2026-05-17-v1","canonical_sha256":"2bb6aaff301bdca42361270cd53b80bb57c27d9af8c34cb262684f4e6a034e51","first_computed_at":"2026-05-18T02:44:15.276775Z","key_id":"pith-v1-2026-05","kind":"pith_receipt","last_reissued_at":"2026-05-18T02:44:15.276775Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","receipt_version":"0.3","signature_b64":"wnaGwgjn2iYg1OWXKIzB4aMguO9VSKWm55q5e/BYGLv6rfIGOR43unAoTRLsR2vViM+ZFFDwVIeoT/3zA0jBCQ==","signature_status":"signed_v1","signed_at":"2026-05-18T02:44:15.277255Z","signed_message":"canonical_sha256_bytes"},"source_id":"2605.13821","source_kind":"arxiv","source_version":1}}},"equivocations":[],"invalid_events":[],"applied_event_ids":["sha256:850d955cf30f7d95940a1e30cf134f5481c401f232d398911d21607ab8002c2b","sha256:df42dc4372c4b225db410d8f805bd581a15c8aa33e60df7e826a099bc298078c"],"state_sha256":"c7ae8de8ec35e158162b9ae6d104f7e726af25aceffaa9ee50ee42c3c1dc095c"}