{"bundle_type":"pith_open_graph_bundle","bundle_version":"1.0","pith_number":"pith:2026:Q6GKIYRWFRYSTWSVFE3UNRYKEX","short_pith_number":"pith:Q6GKIYRW","canonical_record":{"source":{"id":"2605.13296","kind":"arxiv","version":1},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.AI","submitted_at":"2026-05-13T10:10:22Z","cross_cats_sorted":["cs.LG","cs.MA"],"title_canon_sha256":"86da8987df118a32f7575fe174197dd60e3b9dcb1b2ae3ab7b8b848047278994","abstract_canon_sha256":"887e81d9f074ec4f49a9515d145205e84fb52f2de807155aafbf9c97ffb311ad"},"schema_version":"1.0"},"canonical_sha256":"878ca462362c7129da55293746c70a25dce9381ac8fe3d056d2d93c123378229","source":{"kind":"arxiv","id":"2605.13296","version":1},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2605.13296","created_at":"2026-05-18T02:44:49Z"},{"alias_kind":"arxiv_version","alias_value":"2605.13296v1","created_at":"2026-05-18T02:44:49Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2605.13296","created_at":"2026-05-18T02:44:49Z"},{"alias_kind":"pith_short_12","alias_value":"Q6GKIYRWFRYS","created_at":"2026-05-18T12:33:37Z"},{"alias_kind":"pith_short_16","alias_value":"Q6GKIYRWFRYSTWSV","created_at":"2026-05-18T12:33:37Z"},{"alias_kind":"pith_short_8","alias_value":"Q6GKIYRW","created_at":"2026-05-18T12:33:37Z"}],"events":[{"event_type":"record_created","subject_pith_number":"pith:2026:Q6GKIYRWFRYSTWSVFE3UNRYKEX","target":"record","payload":{"canonical_record":{"source":{"id":"2605.13296","kind":"arxiv","version":1},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.AI","submitted_at":"2026-05-13T10:10:22Z","cross_cats_sorted":["cs.LG","cs.MA"],"title_canon_sha256":"86da8987df118a32f7575fe174197dd60e3b9dcb1b2ae3ab7b8b848047278994","abstract_canon_sha256":"887e81d9f074ec4f49a9515d145205e84fb52f2de807155aafbf9c97ffb311ad"},"schema_version":"1.0"},"canonical_sha256":"878ca462362c7129da55293746c70a25dce9381ac8fe3d056d2d93c123378229","receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-05-18T02:44:49.080694Z","signature_b64":"nWLOb9no++ZozteUFr1n7pqYWG6pBtSo8Vu4Fstrc+ICdox+/qyxTr/EZxjZnzpVi/U3NfdmgGRxIY6LrF33Cw==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"878ca462362c7129da55293746c70a25dce9381ac8fe3d056d2d93c123378229","last_reissued_at":"2026-05-18T02:44:49.080261Z","signature_status":"signed_v1","first_computed_at":"2026-05-18T02:44:49.080261Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"source_kind":"arxiv","source_id":"2605.13296","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":"2026-05-18T02:44:49Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"N66B4frvC+umOUrxh4Y3JN3VZwd/3ugCO89li56Hp4sV5v4JpvEjJQaWJivvlF3DxWaYV9dwLlS0C7RWWmbuAA==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-05-21T09:02:59.561382Z"},"content_sha256":"0afb19e0d751bef1d0bdfcc56dcbc41d0d4278a7951d4cd3048e176d198773a9","schema_version":"1.0","event_id":"sha256:0afb19e0d751bef1d0bdfcc56dcbc41d0d4278a7951d4cd3048e176d198773a9"},{"event_type":"graph_snapshot","subject_pith_number":"pith:2026:Q6GKIYRWFRYSTWSVFE3UNRYKEX","target":"graph","payload":{"graph_snapshot":{"paper":{"title":"Discrete Diffusion for Complex and Congested Multi-Agent Path Finding with Sparse Social Attention","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"A discrete diffusion model generates initial joint plans that let a repair solver achieve 95.8 percent success on crowded multi-agent path problems with hundreds of agents.","cross_cats":["cs.LG","cs.MA"],"primary_cat":"cs.AI","authors_text":"Hongguang Wang, Jiaming Guo, Shiyu Quan, Tian Zhi, Xing Hu, Yang Zhao, Yuanzhe Wang, Yunji Chen, Zidong Du, Zihang Wei, Zisheng Liu","submitted_at":"2026-05-13T10:10:22Z","abstract_excerpt":"Multi-Agent Path Finding (MAPF) is a coordination problem that requires computing globally consistent, collision-free trajectories from individual start positions to assigned goal positions under combinatorial planning complexity. In dense environments, suboptimal initial plans induce compound conflicts that hinder feasible repair. For repair-based solvers like LNS2, initial plan quality critically affects downstream repair, yet this factor remains underexplored. We propose DiffLNS, a hybrid framework that integrates a discrete denoising diffusion probabilistic model (D3PM) with LNS2. The D3PM"},"claims":{"count":4,"items":[{"kind":"strongest_claim","text":"Across 20 complex and congested settings, DiffLNS achieves an average success rate of 95.8%, outperforming the strongest tested baseline by 9.6 percentage points and matching or exceeding all baselines in all 20 settings. The initializer generalizes to scenarios with up to 312 agents at inference time despite training only on instances with at most 96 agents.","source":"verdict.strongest_claim","status":"machine_extracted","claim_id":"C1","attestation":"unclaimed"},{"kind":"weakest_assumption","text":"The spatiotemporal prior learned by the discrete diffusion model from expert demonstrations on small instances remains useful and produces repair-friendly drafts when applied to much larger agent counts and unseen congested layouts.","source":"verdict.weakest_assumption","status":"machine_extracted","claim_id":"C2","attestation":"unclaimed"},{"kind":"one_line_summary","text":"DiffLNS uses a discrete diffusion initializer to produce warm-start plans that lift LNS2 success rates to 95.8% across 20 congested MAPF settings, generalizing from 96 to 312 agents.","source":"verdict.one_line_summary","status":"machine_extracted","claim_id":"C3","attestation":"unclaimed"},{"kind":"headline","text":"A discrete diffusion model generates initial joint plans that let a repair solver achieve 95.8 percent success on crowded multi-agent path problems with hundreds of agents.","source":"verdict.pith_extraction.headline","status":"machine_extracted","claim_id":"C4","attestation":"unclaimed"}],"snapshot_sha256":"a000d7fa53440156a83067af7c75487da4f1564b4e52ebe4d84960518e62149b"},"source":{"id":"2605.13296","kind":"arxiv","version":1},"verdict":{"id":"4b3098bf-1dec-4a0f-8f5c-31977f45135f","model_set":{"reader":"grok-4.3"},"created_at":"2026-05-14T19:28:23.982882Z","strongest_claim":"Across 20 complex and congested settings, DiffLNS achieves an average success rate of 95.8%, outperforming the strongest tested baseline by 9.6 percentage points and matching or exceeding all baselines in all 20 settings. The initializer generalizes to scenarios with up to 312 agents at inference time despite training only on instances with at most 96 agents.","one_line_summary":"DiffLNS uses a discrete diffusion initializer to produce warm-start plans that lift LNS2 success rates to 95.8% across 20 congested MAPF settings, generalizing from 96 to 312 agents.","pipeline_version":"pith-pipeline@v0.9.0","weakest_assumption":"The spatiotemporal prior learned by the discrete diffusion model from expert demonstrations on small instances remains useful and produces repair-friendly drafts when applied to much larger agent counts and unseen congested layouts.","pith_extraction_headline":"A discrete diffusion model generates initial joint plans that let a repair solver achieve 95.8 percent success on crowded multi-agent path problems with hundreds of agents."},"references":{"count":26,"sample":[{"doi":"10.1609/aaai.v39i22.34477","year":null,"title":"URL https://ojs.aaai.org/index.php/ AAAI/article/view/34477","work_id":"b354663a-7d91-40fd-9773-8b19a43a395d","ref_index":1,"cited_arxiv_id":"","is_internal_anchor":false},{"doi":"","year":2021,"title":"URL https://proceedings.neurips.cc/paper_files/paper/ 2021/file/958c530554f78bcd8e97125b70e6973d-Paper.pdf. J. Carvalho, A. T. Le, M. Baierl, D. Koert, and J. Peters. Motion planning diffusion: Learni","work_id":"9fdffb25-380a-4c0b-9a49-3dd1652c5666","ref_index":2,"cited_arxiv_id":"","is_internal_anchor":false},{"doi":"","year":1986,"title":"Le, Mark Baierl, Dorothea Koert, and Jan Peters","work_id":"fb3ceb23-685a-4152-be9b-a44af1638a48","ref_index":3,"cited_arxiv_id":"","is_internal_anchor":false},{"doi":"10.1109/robot.1986","year":1986,"title":"doi: 10.1109/ROBOT.1986. 1087401. Emiel Hoogeboom, Didrik Nielsen, Priyank Jaini, Patrick Forré, and Max Welling. Argmax flows and multino- mial diffusion: learning categorical distributions. InProcee","work_id":"b8f49a07-fe85-421e-9648-e7d8bbb182d9","ref_index":4,"cited_arxiv_id":"","is_internal_anchor":false},{"doi":"10.1609/aaai.v36i9.21168","year":2024,"title":"URL https://ojs.aaai.org/index.php/ AAAI/article/view/21168","work_id":"ea666cdf-bb77-4ee7-a37f-ca75529d8292","ref_index":5,"cited_arxiv_id":"","is_internal_anchor":false}],"resolved_work":26,"snapshot_sha256":"45b3b9578c46f444f340b6afd52da11cbe0620f5c18c527d21beb9dbc7466998","internal_anchors":1},"formal_canon":{"evidence_count":2,"snapshot_sha256":"c04985a69e09aca4e3aa595338042afb6da1c3a6b4bb30cc0bc317dc6955b02a"},"author_claims":{"count":0,"strong_count":0,"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"builder_version":"pith-number-builder-2026-05-17-v1"},"verdict_id":"4b3098bf-1dec-4a0f-8f5c-31977f45135f"},"signer":{"signer_id":"pith.science","signer_type":"pith_registry","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"created_at":"2026-05-18T02:44:49Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"xmvaqeDz8zb4t9GGvnRPZEbjPxaOrBIdTgWfftqPeuhA7oTC3f6vqu4w0VX7YKTBf2+DfFpeAlYId/rTzMMaBA==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-05-21T09:02:59.561985Z"},"content_sha256":"723f1cee1827d0266a60bfeef2e7cbc6ed82123edd25409dc3efd633001bf0da","schema_version":"1.0","event_id":"sha256:723f1cee1827d0266a60bfeef2e7cbc6ed82123edd25409dc3efd633001bf0da"},{"event_type":"integrity_finding","subject_pith_number":"pith:2026:Q6GKIYRWFRYSTWSVFE3UNRYKEX","target":"integrity","payload":{"note":"DOI in the printed bibliography is fragmented by whitespace or line breaks. A longer candidate (10.5555/3635637.3663010.URL) was visible in the surrounding text but could not be confirmed against doi.org as printed.","snippet":"doi: 10.5555/3635637.3663010. URL https://dl.acm.org/doi/10. 5555/3635637.3663010. William Peebles and Saining Xie. Scalable diffusion models with transformers. InProceedings of the IEEE/CVF International Conference on Computer Vision (ICCV","arxiv_id":"2605.13296","detector":"doi_compliance","evidence":{"ref_index":11,"verdict_class":"incontrovertible","resolved_title":null,"printed_excerpt":"doi: 10.5555/3635637.3663010. URL https://dl.acm.org/doi/10. 5555/3635637.3663010. William Peebles and Saining Xie. Scalable diffusion models with transformers. InProceedings of the IEEE/CVF International Conference on Computer Vision (ICCV","reconstructed_doi":"10.5555/3635637.3663010.URL"},"severity":"advisory","ref_index":11,"audited_at":"2026-05-19T06:40:27.894528Z","event_type":"pith.integrity.v1","detected_doi":"10.5555/3635637.3663010.URL","detector_url":"https://pith.science/pith-integrity-protocol#doi_compliance","external_url":null,"finding_type":"recoverable_identifier","evidence_hash":"8f59010e469d55b0a4b891c0371f3790ff32be289908abc71d43b9ae98f2b37c","paper_version":1,"verdict_class":"incontrovertible","resolved_title":null,"detector_version":"1.0.0","detected_arxiv_id":null,"integrity_event_id":192,"payload_sha256":"6cf78e5e5cb7b209bdb321320d5836efb17bcfa4053d94f0a6e8f66f3261ee7d","signature_b64":"v6163OLRhRaGni/iPnZ7FDotAN4z0MCs0OpLgZaiC5q6mUosAmtpNOZE/bQhn0pik4golOuf2x/Giv9X6kmjCw==","signing_key_id":"pith-v1-2026-05"},"signer":{"signer_id":"pith.science","signer_type":"pith_registry","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"created_at":"2026-05-19T06:41:53Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"tCPw2vsrevrUe8DfpotTK1igUmQIVlgMqSd/NORu0N46T94rjFeQYUf0Mzu6tNAnQo9qA2eQNaniosfJtGpgBw==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-05-21T09:02:59.562959Z"},"content_sha256":"0ebb1f36f8bbfaf3e982b60bca96b6132ce18ce093e93c12f489f7f8d6f6ac7b","schema_version":"1.0","event_id":"sha256:0ebb1f36f8bbfaf3e982b60bca96b6132ce18ce093e93c12f489f7f8d6f6ac7b"}],"timestamp_proofs":[],"mirror_hints":[{"mirror_type":"https","name":"Pith Resolver","base_url":"https://pith.science","bundle_url":"https://pith.science/pith/Q6GKIYRWFRYSTWSVFE3UNRYKEX/bundle.json","state_url":"https://pith.science/pith/Q6GKIYRWFRYSTWSVFE3UNRYKEX/state.json","well_known_bundle_url":"https://pith.science/.well-known/pith/Q6GKIYRWFRYSTWSVFE3UNRYKEX/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":"2026-05-21T09:02:59Z","links":{"resolver":"https://pith.science/pith/Q6GKIYRWFRYSTWSVFE3UNRYKEX","bundle":"https://pith.science/pith/Q6GKIYRWFRYSTWSVFE3UNRYKEX/bundle.json","state":"https://pith.science/pith/Q6GKIYRWFRYSTWSVFE3UNRYKEX/state.json","well_known_bundle":"https://pith.science/.well-known/pith/Q6GKIYRWFRYSTWSVFE3UNRYKEX/bundle.json"},"state":{"state_type":"pith_open_graph_state","state_version":"1.0","pith_number":"pith:2026:Q6GKIYRWFRYSTWSVFE3UNRYKEX","merge_version":"pith-open-graph-merge-v1","event_count":3,"valid_event_count":3,"invalid_event_count":0,"equivocation_count":0,"current":{"canonical_record":{"metadata":{"abstract_canon_sha256":"887e81d9f074ec4f49a9515d145205e84fb52f2de807155aafbf9c97ffb311ad","cross_cats_sorted":["cs.LG","cs.MA"],"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.AI","submitted_at":"2026-05-13T10:10:22Z","title_canon_sha256":"86da8987df118a32f7575fe174197dd60e3b9dcb1b2ae3ab7b8b848047278994"},"schema_version":"1.0","source":{"id":"2605.13296","kind":"arxiv","version":1}},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2605.13296","created_at":"2026-05-18T02:44:49Z"},{"alias_kind":"arxiv_version","alias_value":"2605.13296v1","created_at":"2026-05-18T02:44:49Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2605.13296","created_at":"2026-05-18T02:44:49Z"},{"alias_kind":"pith_short_12","alias_value":"Q6GKIYRWFRYS","created_at":"2026-05-18T12:33:37Z"},{"alias_kind":"pith_short_16","alias_value":"Q6GKIYRWFRYSTWSV","created_at":"2026-05-18T12:33:37Z"},{"alias_kind":"pith_short_8","alias_value":"Q6GKIYRW","created_at":"2026-05-18T12:33:37Z"}],"graph_snapshots":[{"event_id":"sha256:723f1cee1827d0266a60bfeef2e7cbc6ed82123edd25409dc3efd633001bf0da","target":"graph","created_at":"2026-05-18T02:44:49Z","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":"Across 20 complex and congested settings, DiffLNS achieves an average success rate of 95.8%, outperforming the strongest tested baseline by 9.6 percentage points and matching or exceeding all baselines in all 20 settings. The initializer generalizes to scenarios with up to 312 agents at inference time despite training only on instances with at most 96 agents."},{"attestation":"unclaimed","claim_id":"C2","kind":"weakest_assumption","source":"verdict.weakest_assumption","status":"machine_extracted","text":"The spatiotemporal prior learned by the discrete diffusion model from expert demonstrations on small instances remains useful and produces repair-friendly drafts when applied to much larger agent counts and unseen congested layouts."},{"attestation":"unclaimed","claim_id":"C3","kind":"one_line_summary","source":"verdict.one_line_summary","status":"machine_extracted","text":"DiffLNS uses a discrete diffusion initializer to produce warm-start plans that lift LNS2 success rates to 95.8% across 20 congested MAPF settings, generalizing from 96 to 312 agents."},{"attestation":"unclaimed","claim_id":"C4","kind":"headline","source":"verdict.pith_extraction.headline","status":"machine_extracted","text":"A discrete diffusion model generates initial joint plans that let a repair solver achieve 95.8 percent success on crowded multi-agent path problems with hundreds of agents."}],"snapshot_sha256":"a000d7fa53440156a83067af7c75487da4f1564b4e52ebe4d84960518e62149b"},"formal_canon":{"evidence_count":2,"snapshot_sha256":"c04985a69e09aca4e3aa595338042afb6da1c3a6b4bb30cc0bc317dc6955b02a"},"paper":{"abstract_excerpt":"Multi-Agent Path Finding (MAPF) is a coordination problem that requires computing globally consistent, collision-free trajectories from individual start positions to assigned goal positions under combinatorial planning complexity. In dense environments, suboptimal initial plans induce compound conflicts that hinder feasible repair. For repair-based solvers like LNS2, initial plan quality critically affects downstream repair, yet this factor remains underexplored. We propose DiffLNS, a hybrid framework that integrates a discrete denoising diffusion probabilistic model (D3PM) with LNS2. The D3PM","authors_text":"Hongguang Wang, Jiaming Guo, Shiyu Quan, Tian Zhi, Xing Hu, Yang Zhao, Yuanzhe Wang, Yunji Chen, Zidong Du, Zihang Wei, Zisheng Liu","cross_cats":["cs.LG","cs.MA"],"headline":"A discrete diffusion model generates initial joint plans that let a repair solver achieve 95.8 percent success on crowded multi-agent path problems with hundreds of agents.","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.AI","submitted_at":"2026-05-13T10:10:22Z","title":"Discrete Diffusion for Complex and Congested Multi-Agent Path Finding with Sparse Social Attention"},"references":{"count":26,"internal_anchors":1,"resolved_work":26,"sample":[{"cited_arxiv_id":"","doi":"10.1609/aaai.v39i22.34477","is_internal_anchor":false,"ref_index":1,"title":"URL https://ojs.aaai.org/index.php/ AAAI/article/view/34477","work_id":"b354663a-7d91-40fd-9773-8b19a43a395d","year":null},{"cited_arxiv_id":"","doi":"","is_internal_anchor":false,"ref_index":2,"title":"URL https://proceedings.neurips.cc/paper_files/paper/ 2021/file/958c530554f78bcd8e97125b70e6973d-Paper.pdf. J. Carvalho, A. T. Le, M. Baierl, D. Koert, and J. Peters. Motion planning diffusion: Learni","work_id":"9fdffb25-380a-4c0b-9a49-3dd1652c5666","year":2021},{"cited_arxiv_id":"","doi":"","is_internal_anchor":false,"ref_index":3,"title":"Le, Mark Baierl, Dorothea Koert, and Jan Peters","work_id":"fb3ceb23-685a-4152-be9b-a44af1638a48","year":1986},{"cited_arxiv_id":"","doi":"10.1109/robot.1986","is_internal_anchor":false,"ref_index":4,"title":"doi: 10.1109/ROBOT.1986. 1087401. Emiel Hoogeboom, Didrik Nielsen, Priyank Jaini, Patrick Forré, and Max Welling. Argmax flows and multino- mial diffusion: learning categorical distributions. InProcee","work_id":"b8f49a07-fe85-421e-9648-e7d8bbb182d9","year":1986},{"cited_arxiv_id":"","doi":"10.1609/aaai.v36i9.21168","is_internal_anchor":false,"ref_index":5,"title":"URL https://ojs.aaai.org/index.php/ AAAI/article/view/21168","work_id":"ea666cdf-bb77-4ee7-a37f-ca75529d8292","year":2024}],"snapshot_sha256":"45b3b9578c46f444f340b6afd52da11cbe0620f5c18c527d21beb9dbc7466998"},"source":{"id":"2605.13296","kind":"arxiv","version":1},"verdict":{"created_at":"2026-05-14T19:28:23.982882Z","id":"4b3098bf-1dec-4a0f-8f5c-31977f45135f","model_set":{"reader":"grok-4.3"},"one_line_summary":"DiffLNS uses a discrete diffusion initializer to produce warm-start plans that lift LNS2 success rates to 95.8% across 20 congested MAPF settings, generalizing from 96 to 312 agents.","pipeline_version":"pith-pipeline@v0.9.0","pith_extraction_headline":"A discrete diffusion model generates initial joint plans that let a repair solver achieve 95.8 percent success on crowded multi-agent path problems with hundreds of agents.","strongest_claim":"Across 20 complex and congested settings, DiffLNS achieves an average success rate of 95.8%, outperforming the strongest tested baseline by 9.6 percentage points and matching or exceeding all baselines in all 20 settings. The initializer generalizes to scenarios with up to 312 agents at inference time despite training only on instances with at most 96 agents.","weakest_assumption":"The spatiotemporal prior learned by the discrete diffusion model from expert demonstrations on small instances remains useful and produces repair-friendly drafts when applied to much larger agent counts and unseen congested layouts."}},"verdict_id":"4b3098bf-1dec-4a0f-8f5c-31977f45135f"}}],"author_attestations":[],"timestamp_anchors":[],"storage_attestations":[],"citation_signatures":[],"replication_records":[],"corrections":[],"mirror_hints":[],"record_created":{"event_id":"sha256:0afb19e0d751bef1d0bdfcc56dcbc41d0d4278a7951d4cd3048e176d198773a9","target":"record","created_at":"2026-05-18T02:44:49Z","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":"887e81d9f074ec4f49a9515d145205e84fb52f2de807155aafbf9c97ffb311ad","cross_cats_sorted":["cs.LG","cs.MA"],"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.AI","submitted_at":"2026-05-13T10:10:22Z","title_canon_sha256":"86da8987df118a32f7575fe174197dd60e3b9dcb1b2ae3ab7b8b848047278994"},"schema_version":"1.0","source":{"id":"2605.13296","kind":"arxiv","version":1}},"canonical_sha256":"878ca462362c7129da55293746c70a25dce9381ac8fe3d056d2d93c123378229","receipt":{"algorithm":"ed25519","builder_version":"pith-number-builder-2026-05-17-v1","canonical_sha256":"878ca462362c7129da55293746c70a25dce9381ac8fe3d056d2d93c123378229","first_computed_at":"2026-05-18T02:44:49.080261Z","key_id":"pith-v1-2026-05","kind":"pith_receipt","last_reissued_at":"2026-05-18T02:44:49.080261Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","receipt_version":"0.3","signature_b64":"nWLOb9no++ZozteUFr1n7pqYWG6pBtSo8Vu4Fstrc+ICdox+/qyxTr/EZxjZnzpVi/U3NfdmgGRxIY6LrF33Cw==","signature_status":"signed_v1","signed_at":"2026-05-18T02:44:49.080694Z","signed_message":"canonical_sha256_bytes"},"source_id":"2605.13296","source_kind":"arxiv","source_version":1}}},"equivocations":[],"invalid_events":[],"applied_event_ids":["sha256:0afb19e0d751bef1d0bdfcc56dcbc41d0d4278a7951d4cd3048e176d198773a9","sha256:723f1cee1827d0266a60bfeef2e7cbc6ed82123edd25409dc3efd633001bf0da","sha256:0ebb1f36f8bbfaf3e982b60bca96b6132ce18ce093e93c12f489f7f8d6f6ac7b"],"state_sha256":"54cd7708762e7c78f45c227a0315754d073926387d4016544e33dea0c0287518"},"bundle_signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"zvYSUa2GQc4nguSUk9CLf5lG/lCiSEy+e/diA5H907VkXmoPgTdwlSUqqcNBpx6m4wmism4ToxDHh5V28TrlAw==","signed_message":"bundle_sha256_bytes","signed_at":"2026-05-21T09:02:59.565546Z","bundle_sha256":"7cc16ddefe70ba4bd3d2013a6299cae72eeb85ed18bee9dd259a49ea35f4f736"}}