{"state_type":"pith_open_graph_state","state_version":"1.0","pith_number":"pith:2026:V6Q5QS2RHNDMUQLLJHIQO4IIF2","merge_version":"pith-open-graph-merge-v1","event_count":4,"valid_event_count":4,"invalid_event_count":0,"equivocation_count":1,"current":{"canonical_record":{"metadata":{"abstract_canon_sha256":"c2a12ce8027059bed8a98efe50bb8bc7d2224aa837b8a8ce17bd6d6ff6329a9f","cross_cats_sorted":["cs.DC","cs.MA"],"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.AI","submitted_at":"2026-05-14T17:40:20Z","title_canon_sha256":"89c9c44e7e23810fbdc3af1661355b8d4f65eb58de60ab50ca06d14a1fb92a6e"},"schema_version":"1.0","source":{"id":"2605.15132","kind":"arxiv","version":1}},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2605.15132","created_at":"2026-05-17T21:18:33Z"},{"alias_kind":"arxiv_version","alias_value":"2605.15132v1","created_at":"2026-05-17T21:18:33Z"},{"alias_kind":"pith_short_12","alias_value":"V6Q5QS2RHNDM","created_at":"2026-05-18T12:33:37Z"},{"alias_kind":"pith_short_16","alias_value":"V6Q5QS2RHNDMUQLL","created_at":"2026-05-18T12:33:37Z"},{"alias_kind":"pith_short_8","alias_value":"V6Q5QS2R","created_at":"2026-05-18T12:33:37Z"}],"graph_snapshots":[{"event_id":"sha256:93ac24fb5b01e7bb738660c31b3274f9a97fd43aa1fad1477c42249da4d0ce17","target":"graph","created_at":"2026-05-17T21:57:18Z","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":"In our evaluation, we demonstrate that APWA can dynamically decompose complex queries into parallelizable workflows and scales on larger tasks in settings where prior systems fail completely."},{"attestation":"unclaimed","claim_id":"C2","kind":"weakest_assumption","source":"verdict.weakest_assumption","status":"machine_extracted","text":"That complex agentic workflows can be reliably decomposed into non-interfering subproblems that require no cross-communication and can be executed independently on heterogeneous resources."},{"attestation":"unclaimed","claim_id":"C3","kind":"one_line_summary","source":"verdict.one_line_summary","status":"machine_extracted","text":"APWA is a distributed multi-agent architecture that decomposes parallelizable agentic workflows into non-interfering subproblems for scalable execution on heterogeneous resources."},{"attestation":"unclaimed","claim_id":"C4","kind":"headline","source":"verdict.pith_extraction.headline","status":"machine_extracted","text":"APWA breaks complex agentic tasks into independent subproblems that run in parallel without communication."}],"snapshot_sha256":"2a715faf45f71c4b27f46b68234afa92ccbff58c4ffc9922b56a364cea5020fd"},"formal_canon":{"evidence_count":0,"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"paper":{"abstract_excerpt":"Autonomous multi-agent systems based on large language models (LLMs) have demonstrated remarkable abilities in independently solving complex tasks in a wide breadth of application domains. However, these systems hit critical reasoning, coordination, and computational scaling bottlenecks as the size and complexity of their tasks grow. These limitations hinder multi-agent systems from achieving high-throughput processing for highly parallelizable tasks, despite the availability of parallel computing and reasoning primitives in the underlying LLMs. We introduce the Agent-Parallel Workload Archite","authors_text":"Alina Oprea, Cristina Nita-Rotaru, Evan Rose, Matthew D. Laws, Tushin Mallick","cross_cats":["cs.DC","cs.MA"],"headline":"APWA breaks complex agentic tasks into independent subproblems that run in parallel without communication.","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.AI","submitted_at":"2026-05-14T17:40:20Z","title":"APWA: A Distributed Architecture for Parallelizable Agentic Workflows"},"references":{"count":68,"internal_anchors":0,"resolved_work":68,"sample":[{"cited_arxiv_id":"","doi":"","is_internal_anchor":false,"ref_index":1,"title":"PII masking 300k dataset","work_id":"d7f51784-15a0-4eec-9c05-f5602d7e9177","year":2024},{"cited_arxiv_id":"","doi":"","is_internal_anchor":false,"ref_index":2,"title":"Hadoop.https://hadoop.apache.org/","work_id":"5d5ded8f-4cf6-43fd-85e0-2205545839cd","year":null},{"cited_arxiv_id":"","doi":"","is_internal_anchor":false,"ref_index":3,"title":"Schema-driven information extraction from heterogeneous tables","work_id":"4e98478a-e868-4bc3-bacc-ba1c6bceca41","year":2024},{"cited_arxiv_id":"","doi":"","is_internal_anchor":false,"ref_index":4,"title":"Boiko, Robert MacKnight, Ben Kline, and Gabe Gomes","work_id":"06c818e9-9eb4-4be4-9f23-dba80fb8bf3a","year":2023},{"cited_arxiv_id":"","doi":"","is_internal_anchor":false,"ref_index":5,"title":"Language models are few-shot learners","work_id":"e5d2dfec-d7bd-4be4-9e02-3244247b810a","year":1901}],"snapshot_sha256":"4f1c1ba604e27d6d3d6a0c34b5aa0ac2d8e01d34f45fe34ce41b86ff8a4c4951"},"source":{"id":"2605.15132","kind":"arxiv","version":1},"verdict":{"created_at":"2026-05-15T03:07:47.800391Z","id":"cbd5c22d-b836-4b48-a441-c784b0f8fa2e","model_set":{"reader":"grok-4.3"},"one_line_summary":"APWA is a distributed multi-agent architecture that decomposes parallelizable agentic workflows into non-interfering subproblems for scalable execution on heterogeneous resources.","pipeline_version":"pith-pipeline@v0.9.0","pith_extraction_headline":"APWA breaks complex agentic tasks into independent subproblems that run in parallel without communication.","strongest_claim":"In our evaluation, we demonstrate that APWA can dynamically decompose complex queries into parallelizable workflows and scales on larger tasks in settings where prior systems fail completely.","weakest_assumption":"That complex agentic workflows can be reliably decomposed into non-interfering subproblems that require no cross-communication and can be executed independently on heterogeneous resources."}},"verdict_id":"cbd5c22d-b836-4b48-a441-c784b0f8fa2e"}}],"author_attestations":[],"timestamp_anchors":[],"storage_attestations":[],"citation_signatures":[],"replication_records":[],"corrections":[],"mirror_hints":[],"record_created":{"event_id":"sha256:46da498f24387ca2e6d774d03fe1e2ac01f4d7dcf97193a470ce20811a13ba73","target":"record","created_at":"2026-05-17T21:18:33Z","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":"c2a12ce8027059bed8a98efe50bb8bc7d2224aa837b8a8ce17bd6d6ff6329a9f","cross_cats_sorted":["cs.DC","cs.MA"],"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.AI","submitted_at":"2026-05-14T17:40:20Z","title_canon_sha256":"89c9c44e7e23810fbdc3af1661355b8d4f65eb58de60ab50ca06d14a1fb92a6e"},"schema_version":"1.0","source":{"id":"2605.15132","kind":"arxiv","version":1}},"canonical_sha256":"afa1d84b513b46ca416b49d10771082eadc19a0e0da8e279c418dcb745b885a8","receipt":{"builder_version":"pith-number-builder-2026-05-17-v1","canonical_sha256":"afa1d84b513b46ca416b49d10771082eadc19a0e0da8e279c418dcb745b885a8","first_computed_at":"2026-05-17T21:40:25.620752Z","kind":"pith_receipt","last_reissued_at":"2026-05-17T21:57:18.940110Z","receipt_version":"0.2","signature_status":"unsigned_v0"},"source_id":"2605.15132","source_kind":"arxiv","source_version":1}}},"equivocations":[{"signer_id":"pith.science","event_type":"integrity_finding","target":"integrity","event_ids":["sha256:109911558894708eeb3badf4bf83148232341ec8261c2689f12694b8b6715121","sha256:eb59ff3a26e87bb3e58416e432d6741903e894a7daa391c83d92c111f2f26137"]}],"invalid_events":[],"applied_event_ids":["sha256:46da498f24387ca2e6d774d03fe1e2ac01f4d7dcf97193a470ce20811a13ba73","sha256:93ac24fb5b01e7bb738660c31b3274f9a97fd43aa1fad1477c42249da4d0ce17"],"state_sha256":"bae23bb40684ce36638640cb5709ea7fca0dae920df6b913fd5d7f3f3fc6d4ad"}