{"bundle_type":"pith_open_graph_bundle","bundle_version":"1.0","pith_number":"pith:2025:JIIZUXAAZLOFGFOFXXHBFFKREM","short_pith_number":"pith:JIIZUXAA","canonical_record":{"source":{"id":"2505.03973","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/licenses/by-sa/4.0/","primary_cat":"cs.CL","submitted_at":"2025-05-06T20:50:27Z","cross_cats_sorted":[],"title_canon_sha256":"a923bbed31cae3609cd63b9e0660a6d090673cf051dcd80fe612bb45ac56b9e6","abstract_canon_sha256":"812abfe7fb51ea16f9b6abbdcac075d117459dfa8b88f2a2d8b78e59e38fc5c6"},"schema_version":"1.0"},"canonical_sha256":"4a119a5c00cadc5315c5bdce1295512309bddaff574cd662c14c540f28b49432","source":{"kind":"arxiv","id":"2505.03973","version":1},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2505.03973","created_at":"2026-07-05T10:59:34Z"},{"alias_kind":"arxiv_version","alias_value":"2505.03973v1","created_at":"2026-07-05T10:59:34Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2505.03973","created_at":"2026-07-05T10:59:34Z"},{"alias_kind":"pith_short_12","alias_value":"JIIZUXAAZLOF","created_at":"2026-07-05T10:59:34Z"},{"alias_kind":"pith_short_16","alias_value":"JIIZUXAAZLOFGFOF","created_at":"2026-07-05T10:59:34Z"},{"alias_kind":"pith_short_8","alias_value":"JIIZUXAA","created_at":"2026-07-05T10:59:34Z"}],"events":[{"event_type":"record_created","subject_pith_number":"pith:2025:JIIZUXAAZLOFGFOFXXHBFFKREM","target":"record","payload":{"canonical_record":{"source":{"id":"2505.03973","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/licenses/by-sa/4.0/","primary_cat":"cs.CL","submitted_at":"2025-05-06T20:50:27Z","cross_cats_sorted":[],"title_canon_sha256":"a923bbed31cae3609cd63b9e0660a6d090673cf051dcd80fe612bb45ac56b9e6","abstract_canon_sha256":"812abfe7fb51ea16f9b6abbdcac075d117459dfa8b88f2a2d8b78e59e38fc5c6"},"schema_version":"1.0"},"canonical_sha256":"4a119a5c00cadc5315c5bdce1295512309bddaff574cd662c14c540f28b49432","receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T10:59:34.438520Z","signature_b64":"oiVEXQBzT1/0c4Z387uqNkYMm0vvcauvMrHvROFLlbhhT8/MMokrLnLV6B41MTo4GNfAd1TfHu41woCHS3SiCQ==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"4a119a5c00cadc5315c5bdce1295512309bddaff574cd662c14c540f28b49432","last_reissued_at":"2026-07-05T10:59:34.438106Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T10:59:34.438106Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"source_kind":"arxiv","source_id":"2505.03973","source_version":1,"attestation_state":"computed"},"signer":{"signer_id":"pith.science","signer_type":"pith_registry","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"created_at":"2026-07-05T10:59:34Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"b3AwDsFDHWjFVmiFKbsZFRRsYOEvZj+IhVp5xIZdYsrV8aOLjhVsqo73EQkBF9dvqiBaal+tCAESp0UkI4AUBQ==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-07-10T03:53:12.009173Z"},"content_sha256":"f81d58f369adcdb4324c5eca8cbe64bcc1ae0811995f418f2d698eaada7f191e","schema_version":"1.0","event_id":"sha256:f81d58f369adcdb4324c5eca8cbe64bcc1ae0811995f418f2d698eaada7f191e"},{"event_type":"graph_snapshot","subject_pith_number":"pith:2025:JIIZUXAAZLOFGFOFXXHBFFKREM","target":"graph","payload":{"graph_snapshot":{"paper":{"title":"Divide, Optimize, Merge: Fine-Grained LLM Agent Optimization at Scale","license":"http://creativecommons.org/licenses/by-sa/4.0/","headline":"","cross_cats":[],"primary_cat":"cs.CL","authors_text":"Chi Zhang, Huazheng Wang, Jiale Liu, Malte H{\\o}jmark-Bertelsen, Marie Normann Gadeberg, Qingyun Wu, Shaokun Zhang, Yifan Zeng","submitted_at":"2025-05-06T20:50:27Z","abstract_excerpt":"LLM-based optimization has shown remarkable potential in enhancing agentic systems. However, the conventional approach of prompting LLM optimizer with the whole training trajectories on training dataset in a single pass becomes untenable as datasets grow, leading to context window overflow and degraded pattern recognition. To address these challenges, we propose Fine-Grained Optimization (FGO), a scalable framework that divides large optimization tasks into manageable subsets, performs targeted optimizations, and systematically combines optimized components through progressive merging. Evaluat"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2505.03973","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/2505.03973/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"},"verdict_id":null},"signer":{"signer_id":"pith.science","signer_type":"pith_registry","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"created_at":"2026-07-05T10:59:34Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"nWSiFa5diA8ks06rJjWYrOBfZfQNSiVm6f3kDIbiguGbUlu3M4UaBKHKTvgdQbmdtDKgShfeDz9VIl0yqTbZDA==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-07-10T03:53:12.009634Z"},"content_sha256":"5fedc663d8ff4f2c66e5006fc0f0060891465d4fc9f3cea8a421bbc516c34854","schema_version":"1.0","event_id":"sha256:5fedc663d8ff4f2c66e5006fc0f0060891465d4fc9f3cea8a421bbc516c34854"}],"timestamp_proofs":[],"mirror_hints":[{"mirror_type":"https","name":"Pith Resolver","base_url":"https://pith.science","bundle_url":"https://pith.science/pith/JIIZUXAAZLOFGFOFXXHBFFKREM/bundle.json","state_url":"https://pith.science/pith/JIIZUXAAZLOFGFOFXXHBFFKREM/state.json","well_known_bundle_url":"https://pith.science/.well-known/pith/JIIZUXAAZLOFGFOFXXHBFFKREM/bundle.json","status":"primary"}],"public_keys":[{"key_id":"pith-v1-2026-05","algorithm":"ed25519","format":"raw","public_key_b64":"stVStoiQhXFxp4s2pdzPNoqVNBMojDU/fJ2db5S3CbM=","public_key_hex":"b2d552b68890857171a78b36a5dccf368a953413288c353f7c9d9d6f94b709b3","fingerprint_sha256_b32_first128bits":"RVFV5Z2OI2J3ZUO7ERDEBCYNKS","fingerprint_sha256_hex":"8d4b5ee74e4693bcd1df2446408b0d54","rotates_at":null,"url":"https://pith.science/pith-signing-key.json","notes":"Pith uses this Ed25519 key to sign canonical record SHA-256 digests. Verify with: ed25519_verify(public_key, message=canonical_sha256_bytes, signature=base64decode(signature_b64))."}],"merge_version":"pith-open-graph-merge-v1","built_at":"2026-07-10T03:53:12Z","links":{"resolver":"https://pith.science/pith/JIIZUXAAZLOFGFOFXXHBFFKREM","bundle":"https://pith.science/pith/JIIZUXAAZLOFGFOFXXHBFFKREM/bundle.json","state":"https://pith.science/pith/JIIZUXAAZLOFGFOFXXHBFFKREM/state.json","well_known_bundle":"https://pith.science/.well-known/pith/JIIZUXAAZLOFGFOFXXHBFFKREM/bundle.json"},"state":{"state_type":"pith_open_graph_state","state_version":"1.0","pith_number":"pith:2025:JIIZUXAAZLOFGFOFXXHBFFKREM","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":"812abfe7fb51ea16f9b6abbdcac075d117459dfa8b88f2a2d8b78e59e38fc5c6","cross_cats_sorted":[],"license":"http://creativecommons.org/licenses/by-sa/4.0/","primary_cat":"cs.CL","submitted_at":"2025-05-06T20:50:27Z","title_canon_sha256":"a923bbed31cae3609cd63b9e0660a6d090673cf051dcd80fe612bb45ac56b9e6"},"schema_version":"1.0","source":{"id":"2505.03973","kind":"arxiv","version":1}},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2505.03973","created_at":"2026-07-05T10:59:34Z"},{"alias_kind":"arxiv_version","alias_value":"2505.03973v1","created_at":"2026-07-05T10:59:34Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2505.03973","created_at":"2026-07-05T10:59:34Z"},{"alias_kind":"pith_short_12","alias_value":"JIIZUXAAZLOF","created_at":"2026-07-05T10:59:34Z"},{"alias_kind":"pith_short_16","alias_value":"JIIZUXAAZLOFGFOF","created_at":"2026-07-05T10:59:34Z"},{"alias_kind":"pith_short_8","alias_value":"JIIZUXAA","created_at":"2026-07-05T10:59:34Z"}],"graph_snapshots":[{"event_id":"sha256:5fedc663d8ff4f2c66e5006fc0f0060891465d4fc9f3cea8a421bbc516c34854","target":"graph","created_at":"2026-07-05T10:59:34Z","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":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"formal_canon":{"evidence_count":0,"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"integrity":{"available":true,"clean":true,"detectors_run":[],"endpoint":"/pith/2505.03973/integrity.json","findings":[],"snapshot_sha256":"c28c3603d3b5d939e8dc4c7e95fa8dfce3d595e45f758748cecf8e644a296938","summary":{"advisory":0,"by_detector":{},"critical":0,"informational":0}},"paper":{"abstract_excerpt":"LLM-based optimization has shown remarkable potential in enhancing agentic systems. However, the conventional approach of prompting LLM optimizer with the whole training trajectories on training dataset in a single pass becomes untenable as datasets grow, leading to context window overflow and degraded pattern recognition. To address these challenges, we propose Fine-Grained Optimization (FGO), a scalable framework that divides large optimization tasks into manageable subsets, performs targeted optimizations, and systematically combines optimized components through progressive merging. Evaluat","authors_text":"Chi Zhang, Huazheng Wang, Jiale Liu, Malte H{\\o}jmark-Bertelsen, Marie Normann Gadeberg, Qingyun Wu, Shaokun Zhang, Yifan Zeng","cross_cats":[],"headline":"","license":"http://creativecommons.org/licenses/by-sa/4.0/","primary_cat":"cs.CL","submitted_at":"2025-05-06T20:50:27Z","title":"Divide, Optimize, Merge: Fine-Grained LLM Agent Optimization at Scale"},"references":{"count":0,"internal_anchors":0,"resolved_work":0,"sample":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2505.03973","kind":"arxiv","version":1},"verdict":{"created_at":null,"id":null,"model_set":{},"one_line_summary":"","pipeline_version":null,"pith_extraction_headline":"","strongest_claim":"","weakest_assumption":""}},"verdict_id":null}}],"author_attestations":[],"timestamp_anchors":[],"storage_attestations":[],"citation_signatures":[],"replication_records":[],"corrections":[],"mirror_hints":[],"record_created":{"event_id":"sha256:f81d58f369adcdb4324c5eca8cbe64bcc1ae0811995f418f2d698eaada7f191e","target":"record","created_at":"2026-07-05T10:59:34Z","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":"812abfe7fb51ea16f9b6abbdcac075d117459dfa8b88f2a2d8b78e59e38fc5c6","cross_cats_sorted":[],"license":"http://creativecommons.org/licenses/by-sa/4.0/","primary_cat":"cs.CL","submitted_at":"2025-05-06T20:50:27Z","title_canon_sha256":"a923bbed31cae3609cd63b9e0660a6d090673cf051dcd80fe612bb45ac56b9e6"},"schema_version":"1.0","source":{"id":"2505.03973","kind":"arxiv","version":1}},"canonical_sha256":"4a119a5c00cadc5315c5bdce1295512309bddaff574cd662c14c540f28b49432","receipt":{"algorithm":"ed25519","builder_version":"pith-number-builder-2026-05-17-v1","canonical_sha256":"4a119a5c00cadc5315c5bdce1295512309bddaff574cd662c14c540f28b49432","first_computed_at":"2026-07-05T10:59:34.438106Z","key_id":"pith-v1-2026-05","kind":"pith_receipt","last_reissued_at":"2026-07-05T10:59:34.438106Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","receipt_version":"0.3","signature_b64":"oiVEXQBzT1/0c4Z387uqNkYMm0vvcauvMrHvROFLlbhhT8/MMokrLnLV6B41MTo4GNfAd1TfHu41woCHS3SiCQ==","signature_status":"signed_v1","signed_at":"2026-07-05T10:59:34.438520Z","signed_message":"canonical_sha256_bytes"},"source_id":"2505.03973","source_kind":"arxiv","source_version":1}}},"equivocations":[],"invalid_events":[],"applied_event_ids":["sha256:f81d58f369adcdb4324c5eca8cbe64bcc1ae0811995f418f2d698eaada7f191e","sha256:5fedc663d8ff4f2c66e5006fc0f0060891465d4fc9f3cea8a421bbc516c34854"],"state_sha256":"9752048f10cbc86550fec9ba247edb4b544e2b859a4389d408e9a5171b6fd04a"},"bundle_signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"PYxGT5e3tXzbricTmOT25C1IX06S1SCJD8TgA5lUxSwBfZ3iP0ZH5VpBHq2xON+OxqZ3dc+XFDlCYeHainXTBw==","signed_message":"bundle_sha256_bytes","signed_at":"2026-07-10T03:53:12.011576Z","bundle_sha256":"eeebd09de2123785b813d921e1cd072216321baaad5a77d48624325793aeec0f"}}