{"bundle_type":"pith_open_graph_bundle","bundle_version":"1.0","pith_number":"pith:2026:VGH4GKH4GKCI4ODDSOMW6I3V5C","short_pith_number":"pith:VGH4GKH4","canonical_record":{"source":{"id":"2605.17439","kind":"arxiv","version":1},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.SE","submitted_at":"2026-05-17T13:22:22Z","cross_cats_sorted":["cs.AI"],"title_canon_sha256":"d9eff041f6321d1b73436f9d7179e1a1f664030a60a60027bd085455130cfb58","abstract_canon_sha256":"7b2f0dbb00e3b0e9acb1cb12f2f4f0bda3919442953ac9bb3c084243f8d2a51b"},"schema_version":"1.0"},"canonical_sha256":"a98fc328fc32848e386393996f2375e880d106ea1d335d67386858fdef045956","source":{"kind":"arxiv","id":"2605.17439","version":1},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2605.17439","created_at":"2026-05-20T00:04:38Z"},{"alias_kind":"arxiv_version","alias_value":"2605.17439v1","created_at":"2026-05-20T00:04:38Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2605.17439","created_at":"2026-05-20T00:04:38Z"},{"alias_kind":"pith_short_12","alias_value":"VGH4GKH4GKCI","created_at":"2026-05-20T00:04:38Z"},{"alias_kind":"pith_short_16","alias_value":"VGH4GKH4GKCI4ODD","created_at":"2026-05-20T00:04:38Z"},{"alias_kind":"pith_short_8","alias_value":"VGH4GKH4","created_at":"2026-05-20T00:04:38Z"}],"events":[{"event_type":"record_created","subject_pith_number":"pith:2026:VGH4GKH4GKCI4ODDSOMW6I3V5C","target":"record","payload":{"canonical_record":{"source":{"id":"2605.17439","kind":"arxiv","version":1},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.SE","submitted_at":"2026-05-17T13:22:22Z","cross_cats_sorted":["cs.AI"],"title_canon_sha256":"d9eff041f6321d1b73436f9d7179e1a1f664030a60a60027bd085455130cfb58","abstract_canon_sha256":"7b2f0dbb00e3b0e9acb1cb12f2f4f0bda3919442953ac9bb3c084243f8d2a51b"},"schema_version":"1.0"},"canonical_sha256":"a98fc328fc32848e386393996f2375e880d106ea1d335d67386858fdef045956","receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-05-20T00:04:38.915912Z","signature_b64":"sKDWnljlkBhOMB4Bq+qyJMtE6fx+E8CQNs1EksNKa4af5SrZVrBIg7qMQU+tHEIiXRJUycIPQhkkU6kRkBN6Ag==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"a98fc328fc32848e386393996f2375e880d106ea1d335d67386858fdef045956","last_reissued_at":"2026-05-20T00:04:38.914858Z","signature_status":"signed_v1","first_computed_at":"2026-05-20T00:04:38.914858Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"source_kind":"arxiv","source_id":"2605.17439","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":"2026-05-20T00:04:38Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"7aGkYHmQ3qsEAywrzj53vpeMbTQ15yMTvZWrHqcn2L05tz9X41pEzqINgmQO07tvW3kOqTe9n1npRQuUBMN7Bg==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-05-27T03:03:37.937429Z"},"content_sha256":"3904c239d301f09645f1483e8ddb3b08ecc21e9ebfc88e5637739750c1d3f844","schema_version":"1.0","event_id":"sha256:3904c239d301f09645f1483e8ddb3b08ecc21e9ebfc88e5637739750c1d3f844"},{"event_type":"graph_snapshot","subject_pith_number":"pith:2026:VGH4GKH4GKCI4ODDSOMW6I3V5C","target":"graph","payload":{"graph_snapshot":{"paper":{"title":"DiagEval: Trajectory-Conditioned Diagnosis for Reliable Software Evaluation with GUI Agents","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"Trajectory-conditioned diagnostic probes recover 45-62 percent of failures misattributed to software defects in GUI-agent evaluations.","cross_cats":["cs.AI"],"primary_cat":"cs.SE","authors_text":"Chenglin Wu, Sirui Hong, Tengfei Li, Wei Tao, Yifan Wu, Zhijie Liu","submitted_at":"2026-05-17T13:22:22Z","abstract_excerpt":"Evaluating LLM-generated interactive software requires execution in addition to static analysis. The key difficulty is that correctness is a graph-level reachable property over latent UI state-transition graphs, whereas a GUI evaluator observes only a single execution trajectory. A failed rollout therefore rules out only one realized path, leaving failure attribution ambiguous between evaluator-side execution error and genuine software defect. We present DiagEval, a trajectory-conditioned diagnostic evaluation protocol for post-failure GUI-agent evaluation of interactive software. Rather than "},"claims":{"count":4,"items":[{"kind":"strongest_claim","text":"On false-negative cases, DiagEval recovers 45.6-62.1% of failures that were initially misattributed to software defects, outperforming retry-based baselines with 34.4-160.6% relative gains. On the full evaluation sets, this recovery improves accuracy from 69.9% to 78.3% on WebDevJudge-Unit and from 65.0% to 81.6% on RealDevBench.","source":"verdict.strongest_claim","status":"machine_extracted","claim_id":"C1","attestation":"unclaimed"},{"kind":"weakest_assumption","text":"That targeted diagnostic probes chosen from the failed trajectory can produce an attribution signal that reliably separates evaluator-side execution errors from genuine software defects, without requiring reconstruction of the full latent state-transition graph or calibrated posteriors.","source":"verdict.weakest_assumption","status":"machine_extracted","claim_id":"C2","attestation":"unclaimed"},{"kind":"one_line_summary","text":"DiagEval is a new diagnostic protocol that conditions on failed trajectories to attribute GUI-agent evaluation failures, recovering 45-62% of misattributed cases and lifting accuracy 8-16 points on two benchmarks.","source":"verdict.one_line_summary","status":"machine_extracted","claim_id":"C3","attestation":"unclaimed"},{"kind":"headline","text":"Trajectory-conditioned diagnostic probes recover 45-62 percent of failures misattributed to software defects in GUI-agent evaluations.","source":"verdict.pith_extraction.headline","status":"machine_extracted","claim_id":"C4","attestation":"unclaimed"}],"snapshot_sha256":"afcc07fa1fff788a27cef34eff1a0e62ba51ed66fd51def51068fc3b114f78fd"},"source":{"id":"2605.17439","kind":"arxiv","version":1},"verdict":{"id":"a2ec2395-28e1-45df-a775-ddeadf7c6978","model_set":{"reader":"grok-4.3"},"created_at":"2026-05-19T23:06:42.504265Z","strongest_claim":"On false-negative cases, DiagEval recovers 45.6-62.1% of failures that were initially misattributed to software defects, outperforming retry-based baselines with 34.4-160.6% relative gains. On the full evaluation sets, this recovery improves accuracy from 69.9% to 78.3% on WebDevJudge-Unit and from 65.0% to 81.6% on RealDevBench.","one_line_summary":"DiagEval is a new diagnostic protocol that conditions on failed trajectories to attribute GUI-agent evaluation failures, recovering 45-62% of misattributed cases and lifting accuracy 8-16 points on two benchmarks.","pipeline_version":"pith-pipeline@v0.9.0","weakest_assumption":"That targeted diagnostic probes chosen from the failed trajectory can produce an attribution signal that reliably separates evaluator-side execution errors from genuine software defects, without requiring reconstruction of the full latent state-transition graph or calibrated posteriors.","pith_extraction_headline":"Trajectory-conditioned diagnostic probes recover 45-62 percent of failures misattributed to software defects in GUI-agent evaluations."},"integrity":{"clean":true,"summary":{"advisory":0,"critical":0,"by_detector":{},"informational":0},"endpoint":"/pith/2605.17439/integrity.json","findings":[],"available":true,"detectors_run":[{"name":"doi_title_agreement","ran_at":"2026-05-19T23:31:19.965122Z","status":"completed","version":"1.0.0","findings_count":0},{"name":"doi_compliance","ran_at":"2026-05-19T23:12:26.983482Z","status":"completed","version":"1.0.0","findings_count":0},{"name":"claim_evidence","ran_at":"2026-05-19T21:41:57.723820Z","status":"completed","version":"1.0.0","findings_count":0},{"name":"ai_meta_artifact","ran_at":"2026-05-19T21:33:23.673328Z","status":"skipped","version":"1.0.0","findings_count":0}],"snapshot_sha256":"dda7642886fbcfebc0479a03756fffa82bec8bea606c47fb47d682c6da38af21"},"references":{"count":61,"sample":[{"doi":"","year":null,"title":"Proceedings of the International Conference on Learning Representations (ICLR) , year=","work_id":"6593f727-ae8e-4a7c-a68e-b9dc887a353e","ref_index":1,"cited_arxiv_id":"","is_internal_anchor":false},{"doi":"","year":2024,"title":"Tianbao Xie and Danyang Zhang and Jixuan Chen and Xiaochuan Li and Siheng Zhao and Ruisheng Cao and Toh Jing Hua and Zhoujun Cheng and Dongchan Shin and Fangyu Lei and Yitao Liu and Yiheng Xu and Shuy","work_id":"d87589ab-0362-4d9f-b07d-cbecabe59612","ref_index":2,"cited_arxiv_id":"","is_internal_anchor":false},{"doi":"","year":null,"title":"Proceedings of the International Conference on Learning Representations (ICLR) , year=","work_id":"93bfded3-799c-47e5-9b47-00b19355b539","ref_index":3,"cited_arxiv_id":"","is_internal_anchor":false},{"doi":"","year":2025,"title":"2025 , url =","work_id":"32786012-bfbf-4ee2-bba9-47c2759a57c4","ref_index":4,"cited_arxiv_id":"","is_internal_anchor":false},{"doi":"","year":null,"title":"Advances in Neural Information Processing Systems (NeurIPS) , year=","work_id":"ae7b4e13-4698-4a50-b585-8bbdef9c720c","ref_index":5,"cited_arxiv_id":"","is_internal_anchor":false}],"resolved_work":61,"snapshot_sha256":"a3b9140fe3dbc745651d419e8ce9f7a6c722239d0045eb1159da630eb8d67992","internal_anchors":9},"formal_canon":{"evidence_count":2,"snapshot_sha256":"9a6b5320fdcfabad9f15e4f537cb1bae302ac508e6f3cc0c3b8a55169c5a362c"},"author_claims":{"count":0,"strong_count":0,"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"builder_version":"pith-number-builder-2026-05-17-v1"},"verdict_id":"a2ec2395-28e1-45df-a775-ddeadf7c6978"},"signer":{"signer_id":"pith.science","signer_type":"pith_registry","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"created_at":"2026-05-20T00:04:38Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"HWzeG2AXmGqozfHNDINOhkuy4Kyb9zKFkmSVF9fwql9hGy5kwlJ73UBbxW7HSAuVG7OJzKlFWZia+/DbQf8BDA==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-05-27T03:03:37.938050Z"},"content_sha256":"5f60e4d4ae65864b31d018f60b545e0cdf63e9cdabc2615f2176a32cf1915bd4","schema_version":"1.0","event_id":"sha256:5f60e4d4ae65864b31d018f60b545e0cdf63e9cdabc2615f2176a32cf1915bd4"}],"timestamp_proofs":[],"mirror_hints":[{"mirror_type":"https","name":"Pith Resolver","base_url":"https://pith.science","bundle_url":"https://pith.science/pith/VGH4GKH4GKCI4ODDSOMW6I3V5C/bundle.json","state_url":"https://pith.science/pith/VGH4GKH4GKCI4ODDSOMW6I3V5C/state.json","well_known_bundle_url":"https://pith.science/.well-known/pith/VGH4GKH4GKCI4ODDSOMW6I3V5C/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":"2026-05-27T03:03:37Z","links":{"resolver":"https://pith.science/pith/VGH4GKH4GKCI4ODDSOMW6I3V5C","bundle":"https://pith.science/pith/VGH4GKH4GKCI4ODDSOMW6I3V5C/bundle.json","state":"https://pith.science/pith/VGH4GKH4GKCI4ODDSOMW6I3V5C/state.json","well_known_bundle":"https://pith.science/.well-known/pith/VGH4GKH4GKCI4ODDSOMW6I3V5C/bundle.json"},"state":{"state_type":"pith_open_graph_state","state_version":"1.0","pith_number":"pith:2026:VGH4GKH4GKCI4ODDSOMW6I3V5C","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":"7b2f0dbb00e3b0e9acb1cb12f2f4f0bda3919442953ac9bb3c084243f8d2a51b","cross_cats_sorted":["cs.AI"],"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.SE","submitted_at":"2026-05-17T13:22:22Z","title_canon_sha256":"d9eff041f6321d1b73436f9d7179e1a1f664030a60a60027bd085455130cfb58"},"schema_version":"1.0","source":{"id":"2605.17439","kind":"arxiv","version":1}},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2605.17439","created_at":"2026-05-20T00:04:38Z"},{"alias_kind":"arxiv_version","alias_value":"2605.17439v1","created_at":"2026-05-20T00:04:38Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2605.17439","created_at":"2026-05-20T00:04:38Z"},{"alias_kind":"pith_short_12","alias_value":"VGH4GKH4GKCI","created_at":"2026-05-20T00:04:38Z"},{"alias_kind":"pith_short_16","alias_value":"VGH4GKH4GKCI4ODD","created_at":"2026-05-20T00:04:38Z"},{"alias_kind":"pith_short_8","alias_value":"VGH4GKH4","created_at":"2026-05-20T00:04:38Z"}],"graph_snapshots":[{"event_id":"sha256:5f60e4d4ae65864b31d018f60b545e0cdf63e9cdabc2615f2176a32cf1915bd4","target":"graph","created_at":"2026-05-20T00:04:38Z","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":"On false-negative cases, DiagEval recovers 45.6-62.1% of failures that were initially misattributed to software defects, outperforming retry-based baselines with 34.4-160.6% relative gains. On the full evaluation sets, this recovery improves accuracy from 69.9% to 78.3% on WebDevJudge-Unit and from 65.0% to 81.6% on RealDevBench."},{"attestation":"unclaimed","claim_id":"C2","kind":"weakest_assumption","source":"verdict.weakest_assumption","status":"machine_extracted","text":"That targeted diagnostic probes chosen from the failed trajectory can produce an attribution signal that reliably separates evaluator-side execution errors from genuine software defects, without requiring reconstruction of the full latent state-transition graph or calibrated posteriors."},{"attestation":"unclaimed","claim_id":"C3","kind":"one_line_summary","source":"verdict.one_line_summary","status":"machine_extracted","text":"DiagEval is a new diagnostic protocol that conditions on failed trajectories to attribute GUI-agent evaluation failures, recovering 45-62% of misattributed cases and lifting accuracy 8-16 points on two benchmarks."},{"attestation":"unclaimed","claim_id":"C4","kind":"headline","source":"verdict.pith_extraction.headline","status":"machine_extracted","text":"Trajectory-conditioned diagnostic probes recover 45-62 percent of failures misattributed to software defects in GUI-agent evaluations."}],"snapshot_sha256":"afcc07fa1fff788a27cef34eff1a0e62ba51ed66fd51def51068fc3b114f78fd"},"formal_canon":{"evidence_count":2,"snapshot_sha256":"9a6b5320fdcfabad9f15e4f537cb1bae302ac508e6f3cc0c3b8a55169c5a362c"},"integrity":{"available":true,"clean":true,"detectors_run":[{"findings_count":0,"name":"doi_title_agreement","ran_at":"2026-05-19T23:31:19.965122Z","status":"completed","version":"1.0.0"},{"findings_count":0,"name":"doi_compliance","ran_at":"2026-05-19T23:12:26.983482Z","status":"completed","version":"1.0.0"},{"findings_count":0,"name":"claim_evidence","ran_at":"2026-05-19T21:41:57.723820Z","status":"completed","version":"1.0.0"},{"findings_count":0,"name":"ai_meta_artifact","ran_at":"2026-05-19T21:33:23.673328Z","status":"skipped","version":"1.0.0"}],"endpoint":"/pith/2605.17439/integrity.json","findings":[],"snapshot_sha256":"dda7642886fbcfebc0479a03756fffa82bec8bea606c47fb47d682c6da38af21","summary":{"advisory":0,"by_detector":{},"critical":0,"informational":0}},"paper":{"abstract_excerpt":"Evaluating LLM-generated interactive software requires execution in addition to static analysis. The key difficulty is that correctness is a graph-level reachable property over latent UI state-transition graphs, whereas a GUI evaluator observes only a single execution trajectory. A failed rollout therefore rules out only one realized path, leaving failure attribution ambiguous between evaluator-side execution error and genuine software defect. We present DiagEval, a trajectory-conditioned diagnostic evaluation protocol for post-failure GUI-agent evaluation of interactive software. Rather than ","authors_text":"Chenglin Wu, Sirui Hong, Tengfei Li, Wei Tao, Yifan Wu, Zhijie Liu","cross_cats":["cs.AI"],"headline":"Trajectory-conditioned diagnostic probes recover 45-62 percent of failures misattributed to software defects in GUI-agent evaluations.","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.SE","submitted_at":"2026-05-17T13:22:22Z","title":"DiagEval: Trajectory-Conditioned Diagnosis for Reliable Software Evaluation with GUI Agents"},"references":{"count":61,"internal_anchors":9,"resolved_work":61,"sample":[{"cited_arxiv_id":"","doi":"","is_internal_anchor":false,"ref_index":1,"title":"Proceedings of the International Conference on Learning Representations (ICLR) , year=","work_id":"6593f727-ae8e-4a7c-a68e-b9dc887a353e","year":null},{"cited_arxiv_id":"","doi":"","is_internal_anchor":false,"ref_index":2,"title":"Tianbao Xie and Danyang Zhang and Jixuan Chen and Xiaochuan Li and Siheng Zhao and Ruisheng Cao and Toh Jing Hua and Zhoujun Cheng and Dongchan Shin and Fangyu Lei and Yitao Liu and Yiheng Xu and Shuy","work_id":"d87589ab-0362-4d9f-b07d-cbecabe59612","year":2024},{"cited_arxiv_id":"","doi":"","is_internal_anchor":false,"ref_index":3,"title":"Proceedings of the International Conference on Learning Representations (ICLR) , year=","work_id":"93bfded3-799c-47e5-9b47-00b19355b539","year":null},{"cited_arxiv_id":"","doi":"","is_internal_anchor":false,"ref_index":4,"title":"2025 , url =","work_id":"32786012-bfbf-4ee2-bba9-47c2759a57c4","year":2025},{"cited_arxiv_id":"","doi":"","is_internal_anchor":false,"ref_index":5,"title":"Advances in Neural Information Processing Systems (NeurIPS) , year=","work_id":"ae7b4e13-4698-4a50-b585-8bbdef9c720c","year":null}],"snapshot_sha256":"a3b9140fe3dbc745651d419e8ce9f7a6c722239d0045eb1159da630eb8d67992"},"source":{"id":"2605.17439","kind":"arxiv","version":1},"verdict":{"created_at":"2026-05-19T23:06:42.504265Z","id":"a2ec2395-28e1-45df-a775-ddeadf7c6978","model_set":{"reader":"grok-4.3"},"one_line_summary":"DiagEval is a new diagnostic protocol that conditions on failed trajectories to attribute GUI-agent evaluation failures, recovering 45-62% of misattributed cases and lifting accuracy 8-16 points on two benchmarks.","pipeline_version":"pith-pipeline@v0.9.0","pith_extraction_headline":"Trajectory-conditioned diagnostic probes recover 45-62 percent of failures misattributed to software defects in GUI-agent evaluations.","strongest_claim":"On false-negative cases, DiagEval recovers 45.6-62.1% of failures that were initially misattributed to software defects, outperforming retry-based baselines with 34.4-160.6% relative gains. On the full evaluation sets, this recovery improves accuracy from 69.9% to 78.3% on WebDevJudge-Unit and from 65.0% to 81.6% on RealDevBench.","weakest_assumption":"That targeted diagnostic probes chosen from the failed trajectory can produce an attribution signal that reliably separates evaluator-side execution errors from genuine software defects, without requiring reconstruction of the full latent state-transition graph or calibrated posteriors."}},"verdict_id":"a2ec2395-28e1-45df-a775-ddeadf7c6978"}}],"author_attestations":[],"timestamp_anchors":[],"storage_attestations":[],"citation_signatures":[],"replication_records":[],"corrections":[],"mirror_hints":[],"record_created":{"event_id":"sha256:3904c239d301f09645f1483e8ddb3b08ecc21e9ebfc88e5637739750c1d3f844","target":"record","created_at":"2026-05-20T00:04:38Z","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":"7b2f0dbb00e3b0e9acb1cb12f2f4f0bda3919442953ac9bb3c084243f8d2a51b","cross_cats_sorted":["cs.AI"],"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.SE","submitted_at":"2026-05-17T13:22:22Z","title_canon_sha256":"d9eff041f6321d1b73436f9d7179e1a1f664030a60a60027bd085455130cfb58"},"schema_version":"1.0","source":{"id":"2605.17439","kind":"arxiv","version":1}},"canonical_sha256":"a98fc328fc32848e386393996f2375e880d106ea1d335d67386858fdef045956","receipt":{"algorithm":"ed25519","builder_version":"pith-number-builder-2026-05-17-v1","canonical_sha256":"a98fc328fc32848e386393996f2375e880d106ea1d335d67386858fdef045956","first_computed_at":"2026-05-20T00:04:38.914858Z","key_id":"pith-v1-2026-05","kind":"pith_receipt","last_reissued_at":"2026-05-20T00:04:38.914858Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","receipt_version":"0.3","signature_b64":"sKDWnljlkBhOMB4Bq+qyJMtE6fx+E8CQNs1EksNKa4af5SrZVrBIg7qMQU+tHEIiXRJUycIPQhkkU6kRkBN6Ag==","signature_status":"signed_v1","signed_at":"2026-05-20T00:04:38.915912Z","signed_message":"canonical_sha256_bytes"},"source_id":"2605.17439","source_kind":"arxiv","source_version":1}}},"equivocations":[],"invalid_events":[],"applied_event_ids":["sha256:3904c239d301f09645f1483e8ddb3b08ecc21e9ebfc88e5637739750c1d3f844","sha256:5f60e4d4ae65864b31d018f60b545e0cdf63e9cdabc2615f2176a32cf1915bd4"],"state_sha256":"8ce30aa6fbfe8b7d7a74f542524cb7687ff424e761b47d6c4629e7a39b666ed4"},"bundle_signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"R7xylwg+XFonLL0nwPzz1UevCHrvYXOuhBzmEASubzeDUZPbY2F0RCTZhHcp1UP3NAXni82AxR1lsqYh6vLFAQ==","signed_message":"bundle_sha256_bytes","signed_at":"2026-05-27T03:03:37.942134Z","bundle_sha256":"f0d30380b5d7c0c901e87a0903f1556cd1a4bec78dfd1dfab67fdc83c1adcbdf"}}