{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2024:JKL3JYFNJYUPZPFLAUJUKL762V","short_pith_number":"pith:JKL3JYFN","schema_version":"1.0","canonical_sha256":"4a97b4e0ad4e28fcbcab0513452ffed55c4b93a1c40f8d4dff43c8c04862ed00","source":{"kind":"arxiv","id":"2402.16823","version":3},"attestation_state":"computed","paper":{"title":"Language Agents as Optimizable Graphs","license":"http://creativecommons.org/licenses/by-nc-sa/4.0/","headline":"","cross_cats":["cs.CL","cs.LG","cs.MA"],"primary_cat":"cs.AI","authors_text":"Dmitrii Khizbullin, Francesco Faccio, J\\\"urgen Schmidhuber, Louis Kirsch, Mingchen Zhuge, Wenyi Wang","submitted_at":"2024-02-26T18:48:27Z","abstract_excerpt":"Various human-designed prompt engineering techniques have been proposed to improve problem solvers based on Large Language Models (LLMs), yielding many disparate code bases. We unify these approaches by describing LLM-based agents as computational graphs. The nodes implement functions to process multimodal data or query LLMs, and the edges describe the information flow between operations. Graphs can be recursively combined into larger composite graphs representing hierarchies of inter-agent collaboration (where edges connect operations of different agents). Our novel automatic graph optimizers"},"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":"2402.16823","kind":"arxiv","version":3},"metadata":{"license":"http://creativecommons.org/licenses/by-nc-sa/4.0/","primary_cat":"cs.AI","submitted_at":"2024-02-26T18:48:27Z","cross_cats_sorted":["cs.CL","cs.LG","cs.MA"],"title_canon_sha256":"8e4e27966f221c8667122fba7282bc48e45a4506584aeff1fa5b6f28650ab71d","abstract_canon_sha256":"86291045ee1344db338f9cbed50c054f29ff5bfdf459398e11d2953ff4eeda8f"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T08:57:55.089665Z","signature_b64":"DXd9M7GjH3ZB44MJts4wXHNgdDet6lMpvV+0Mj+zXmoVgCpXrKTbTT0FIOR3sl6Wc7P+02OSMHaky/qGY8UoAw==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"4a97b4e0ad4e28fcbcab0513452ffed55c4b93a1c40f8d4dff43c8c04862ed00","last_reissued_at":"2026-07-05T08:57:55.089153Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T08:57:55.089153Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Language Agents as Optimizable Graphs","license":"http://creativecommons.org/licenses/by-nc-sa/4.0/","headline":"","cross_cats":["cs.CL","cs.LG","cs.MA"],"primary_cat":"cs.AI","authors_text":"Dmitrii Khizbullin, Francesco Faccio, J\\\"urgen Schmidhuber, Louis Kirsch, Mingchen Zhuge, Wenyi Wang","submitted_at":"2024-02-26T18:48:27Z","abstract_excerpt":"Various human-designed prompt engineering techniques have been proposed to improve problem solvers based on Large Language Models (LLMs), yielding many disparate code bases. We unify these approaches by describing LLM-based agents as computational graphs. The nodes implement functions to process multimodal data or query LLMs, and the edges describe the information flow between operations. Graphs can be recursively combined into larger composite graphs representing hierarchies of inter-agent collaboration (where edges connect operations of different agents). Our novel automatic graph optimizers"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2402.16823","kind":"arxiv","version":3},"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/2402.16823/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":"2402.16823","created_at":"2026-07-05T08:57:55.089211+00:00"},{"alias_kind":"arxiv_version","alias_value":"2402.16823v3","created_at":"2026-07-05T08:57:55.089211+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2402.16823","created_at":"2026-07-05T08:57:55.089211+00:00"},{"alias_kind":"pith_short_12","alias_value":"JKL3JYFNJYUP","created_at":"2026-07-05T08:57:55.089211+00:00"},{"alias_kind":"pith_short_16","alias_value":"JKL3JYFNJYUPZPFL","created_at":"2026-07-05T08:57:55.089211+00:00"},{"alias_kind":"pith_short_8","alias_value":"JKL3JYFN","created_at":"2026-07-05T08:57:55.089211+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":13,"internal_anchor_count":0,"sample":[{"citing_arxiv_id":"2606.10402","citing_title":"Harnessing the Collective Intelligence of AI Agents in the Wild for New Discoveries","ref_index":39,"is_internal_anchor":false},{"citing_arxiv_id":"2606.09498","citing_title":"Self-Harness: Harnesses That Improve Themselves","ref_index":37,"is_internal_anchor":false},{"citing_arxiv_id":"2606.30602","citing_title":"MESA: Prioritizing Vulnerable Communication Channels for Securing Multi-Agent Systems","ref_index":29,"is_internal_anchor":false},{"citing_arxiv_id":"2605.28655","citing_title":"AutoScientists: Self-Organizing Agent Teams for Long-Running Scientific Experimentation","ref_index":37,"is_internal_anchor":false},{"citing_arxiv_id":"2310.02170","citing_title":"A Dynamic LLM-Powered Agent Network for Task-Oriented Agent Collaboration","ref_index":34,"is_internal_anchor":false},{"citing_arxiv_id":"2510.07799","citing_title":"Dynamic Generation of Multi-LLM Agents Communication Topologies with Graph Diffusion Models","ref_index":38,"is_internal_anchor":false},{"citing_arxiv_id":"2605.20425","citing_title":"AgentCo-op: Retrieval-Based Synthesis of Interoperable Multi-Agent Workflows","ref_index":31,"is_internal_anchor":false},{"citing_arxiv_id":"2604.09611","citing_title":"Characterizing Performance-Energy Trade-offs of Large Language Models in Multi-Request Workflows","ref_index":78,"is_internal_anchor":false},{"citing_arxiv_id":"2605.08715","citing_title":"AgentForesight: Online Auditing for Early Failure Prediction in Multi-Agent Systems","ref_index":70,"is_internal_anchor":false},{"citing_arxiv_id":"2605.08715","citing_title":"AgentForesight: Online Auditing for Early Failure Prediction in Multi-Agent Systems","ref_index":70,"is_internal_anchor":false},{"citing_arxiv_id":"2605.05657","citing_title":"Retrieval-Conditioned Topology Selection with Provable Budget Conservation for Multi-Agent Code Generation","ref_index":106,"is_internal_anchor":false},{"citing_arxiv_id":"2605.07073","citing_title":"TeamBench: Evaluating Agent Coordination under Enforced Role Separation","ref_index":5,"is_internal_anchor":false},{"citing_arxiv_id":"2604.06452","citing_title":"Learning to Interrupt in Language-based Multi-agent Communication","ref_index":48,"is_internal_anchor":false}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/JKL3JYFNJYUPZPFLAUJUKL762V","json":"https://pith.science/pith/JKL3JYFNJYUPZPFLAUJUKL762V.json","graph_json":"https://pith.science/api/pith-number/JKL3JYFNJYUPZPFLAUJUKL762V/graph.json","events_json":"https://pith.science/api/pith-number/JKL3JYFNJYUPZPFLAUJUKL762V/events.json","paper":"https://pith.science/paper/JKL3JYFN"},"agent_actions":{"view_html":"https://pith.science/pith/JKL3JYFNJYUPZPFLAUJUKL762V","download_json":"https://pith.science/pith/JKL3JYFNJYUPZPFLAUJUKL762V.json","view_paper":"https://pith.science/paper/JKL3JYFN","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2402.16823&json=true","fetch_graph":"https://pith.science/api/pith-number/JKL3JYFNJYUPZPFLAUJUKL762V/graph.json","fetch_events":"https://pith.science/api/pith-number/JKL3JYFNJYUPZPFLAUJUKL762V/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/JKL3JYFNJYUPZPFLAUJUKL762V/action/timestamp_anchor","attest_storage":"https://pith.science/pith/JKL3JYFNJYUPZPFLAUJUKL762V/action/storage_attestation","attest_author":"https://pith.science/pith/JKL3JYFNJYUPZPFLAUJUKL762V/action/author_attestation","sign_citation":"https://pith.science/pith/JKL3JYFNJYUPZPFLAUJUKL762V/action/citation_signature","submit_replication":"https://pith.science/pith/JKL3JYFNJYUPZPFLAUJUKL762V/action/replication_record"}},"created_at":"2026-07-05T08:57:55.089211+00:00","updated_at":"2026-07-05T08:57:55.089211+00:00"}