{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2026:YNLJFTQMX722DJM6XCMI2DM7YQ","short_pith_number":"pith:YNLJFTQM","schema_version":"1.0","canonical_sha256":"c35692ce0cbff5a1a59eb8988d0d9fc40ec8825bad177582310c07a4559feb33","source":{"kind":"arxiv","id":"2606.03099","version":1},"attestation_state":"computed","paper":{"title":"PhotoCraft: Agentic Reasoning with Hierarchical Self-Evolving Memory for Deep Image Search","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.AI"],"primary_cat":"cs.CL","authors_text":"Ce Hao, Jianwei He, Jie Zhou, Jinchao Zhang, Kailin Lyu, Lianyu Hu, Nanxing Hu, Qiwei Yan, Shengqian Qin, Xuanbo Su, Yang Liu, Zhiqiang Yuan","submitted_at":"2026-06-02T03:38:44Z","abstract_excerpt":"Deep Image Search requires multi-step reasoning over rich contextual cues, such as time, location, and event relations. However, most existing LLM-based agents are stateless and reactive, lacking persistent memory to maintain long-horizon context or transfer experience across tasks, which often leads to execution drift and experience isolation. To address these limitations, we propose PhotoCraft, a training-free, hierarchical memory system for photo-search agents. Inspired by human cognition, PhotoCraft equips MLLMs with working, episodic, and semantic memory, which are dynamically invoked dur"},"verification_status":{"content_addressed":true,"pith_receipt":true,"author_attested":false,"weak_author_claims":0,"strong_author_claims":0,"externally_anchored":false,"storage_verified":false,"citation_signatures":0,"replication_records":0,"graph_snapshot":true,"references_resolved":false,"formal_links_present":false},"canonical_record":{"source":{"id":"2606.03099","kind":"arxiv","version":1},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CL","submitted_at":"2026-06-02T03:38:44Z","cross_cats_sorted":["cs.AI"],"title_canon_sha256":"08726578de5c3d0b3fa64c4be1fe79d6d40852bc77cf2fbeda68c99f832dff0e","abstract_canon_sha256":"8395039fba3cf744186cc58d12ebbc01363795be3f4101996891704afabe59da"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-06-03T01:05:31.673530Z","signature_b64":"xxgL2Ya9/w14+Et1SRPKklEgndNUyqoBBQ4AbL7NvhTtOXsPFQ9YQxMhqICsqhbaNBywdX2zldWXvBc84Xy9AA==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"c35692ce0cbff5a1a59eb8988d0d9fc40ec8825bad177582310c07a4559feb33","last_reissued_at":"2026-06-03T01:05:31.673120Z","signature_status":"signed_v1","first_computed_at":"2026-06-03T01:05:31.673120Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"PhotoCraft: Agentic Reasoning with Hierarchical Self-Evolving Memory for Deep Image Search","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.AI"],"primary_cat":"cs.CL","authors_text":"Ce Hao, Jianwei He, Jie Zhou, Jinchao Zhang, Kailin Lyu, Lianyu Hu, Nanxing Hu, Qiwei Yan, Shengqian Qin, Xuanbo Su, Yang Liu, Zhiqiang Yuan","submitted_at":"2026-06-02T03:38:44Z","abstract_excerpt":"Deep Image Search requires multi-step reasoning over rich contextual cues, such as time, location, and event relations. However, most existing LLM-based agents are stateless and reactive, lacking persistent memory to maintain long-horizon context or transfer experience across tasks, which often leads to execution drift and experience isolation. To address these limitations, we propose PhotoCraft, a training-free, hierarchical memory system for photo-search agents. Inspired by human cognition, PhotoCraft equips MLLMs with working, episodic, and semantic memory, which are dynamically invoked dur"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2606.03099","kind":"arxiv","version":1},"verdict":{"id":null,"model_set":{},"created_at":null,"strongest_claim":"","one_line_summary":"","pipeline_version":null,"weakest_assumption":"","pith_extraction_headline":""},"integrity":{"clean":true,"summary":{"advisory":0,"critical":0,"by_detector":{},"informational":0},"endpoint":"/pith/2606.03099/integrity.json","findings":[],"available":true,"detectors_run":[],"snapshot_sha256":"c28c3603d3b5d939e8dc4c7e95fa8dfce3d595e45f758748cecf8e644a296938"},"references":{"count":0,"sample":[],"resolved_work":0,"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57","internal_anchors":0},"formal_canon":{"evidence_count":0,"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"author_claims":{"count":0,"strong_count":0,"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"builder_version":"pith-number-builder-2026-05-17-v1"},"aliases":[{"alias_kind":"arxiv","alias_value":"2606.03099","created_at":"2026-06-03T01:05:31.673166+00:00"},{"alias_kind":"arxiv_version","alias_value":"2606.03099v1","created_at":"2026-06-03T01:05:31.673166+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2606.03099","created_at":"2026-06-03T01:05:31.673166+00:00"},{"alias_kind":"pith_short_12","alias_value":"YNLJFTQMX722","created_at":"2026-06-03T01:05:31.673166+00:00"},{"alias_kind":"pith_short_16","alias_value":"YNLJFTQMX722DJM6","created_at":"2026-06-03T01:05:31.673166+00:00"},{"alias_kind":"pith_short_8","alias_value":"YNLJFTQM","created_at":"2026-06-03T01:05:31.673166+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":0,"internal_anchor_count":0,"sample":[]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/YNLJFTQMX722DJM6XCMI2DM7YQ","json":"https://pith.science/pith/YNLJFTQMX722DJM6XCMI2DM7YQ.json","graph_json":"https://pith.science/api/pith-number/YNLJFTQMX722DJM6XCMI2DM7YQ/graph.json","events_json":"https://pith.science/api/pith-number/YNLJFTQMX722DJM6XCMI2DM7YQ/events.json","paper":"https://pith.science/paper/YNLJFTQM"},"agent_actions":{"view_html":"https://pith.science/pith/YNLJFTQMX722DJM6XCMI2DM7YQ","download_json":"https://pith.science/pith/YNLJFTQMX722DJM6XCMI2DM7YQ.json","view_paper":"https://pith.science/paper/YNLJFTQM","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2606.03099&json=true","fetch_graph":"https://pith.science/api/pith-number/YNLJFTQMX722DJM6XCMI2DM7YQ/graph.json","fetch_events":"https://pith.science/api/pith-number/YNLJFTQMX722DJM6XCMI2DM7YQ/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/YNLJFTQMX722DJM6XCMI2DM7YQ/action/timestamp_anchor","attest_storage":"https://pith.science/pith/YNLJFTQMX722DJM6XCMI2DM7YQ/action/storage_attestation","attest_author":"https://pith.science/pith/YNLJFTQMX722DJM6XCMI2DM7YQ/action/author_attestation","sign_citation":"https://pith.science/pith/YNLJFTQMX722DJM6XCMI2DM7YQ/action/citation_signature","submit_replication":"https://pith.science/pith/YNLJFTQMX722DJM6XCMI2DM7YQ/action/replication_record"}},"created_at":"2026-06-03T01:05:31.673166+00:00","updated_at":"2026-06-03T01:05:31.673166+00:00"}