{"bundle_type":"pith_open_graph_bundle","bundle_version":"1.0","pith_number":"pith:2026:WNSH2SFNBIHTIOCFFSFHULXAZQ","short_pith_number":"pith:WNSH2SFN","canonical_record":{"source":{"id":"2607.27283","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.LG","submitted_at":"2026-07-29T13:57:32Z","cross_cats_sorted":["cs.AI","cs.SE"],"title_canon_sha256":"7ef9f2c8fcd9a29097520cf3d7289d4215c7772b64bceead9eb55a677035602b","abstract_canon_sha256":"3c8dad46a4fff4e8c3b267cf261b5f46815da42d157decbd67802bd6c14d1ca1"},"schema_version":"1.0"},"canonical_sha256":"b3647d48ad0a0f3438452c8a7a2ee0cc2914f610f48cfbf1b6bb2dcf162418d4","source":{"kind":"arxiv","id":"2607.27283","version":1},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2607.27283","created_at":"2026-07-31T00:10:29Z"},{"alias_kind":"arxiv_version","alias_value":"2607.27283v1","created_at":"2026-07-31T00:10:29Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2607.27283","created_at":"2026-07-31T00:10:29Z"},{"alias_kind":"pith_short_12","alias_value":"WNSH2SFNBIHT","created_at":"2026-07-31T00:10:29Z"},{"alias_kind":"pith_short_16","alias_value":"WNSH2SFNBIHTIOCF","created_at":"2026-07-31T00:10:29Z"},{"alias_kind":"pith_short_8","alias_value":"WNSH2SFN","created_at":"2026-07-31T00:10:29Z"}],"events":[{"event_type":"record_created","subject_pith_number":"pith:2026:WNSH2SFNBIHTIOCFFSFHULXAZQ","target":"record","payload":{"canonical_record":{"source":{"id":"2607.27283","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.LG","submitted_at":"2026-07-29T13:57:32Z","cross_cats_sorted":["cs.AI","cs.SE"],"title_canon_sha256":"7ef9f2c8fcd9a29097520cf3d7289d4215c7772b64bceead9eb55a677035602b","abstract_canon_sha256":"3c8dad46a4fff4e8c3b267cf261b5f46815da42d157decbd67802bd6c14d1ca1"},"schema_version":"1.0"},"canonical_sha256":"b3647d48ad0a0f3438452c8a7a2ee0cc2914f610f48cfbf1b6bb2dcf162418d4","receipt":{"kind":"pith_receipt","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"b3647d48ad0a0f3438452c8a7a2ee0cc2914f610f48cfbf1b6bb2dcf162418d4","last_reissued_at":"2026-07-31T00:10:29.953952Z","signature_status":"unsigned_v0","first_computed_at":"2026-07-31T00:10:29.953952Z"},"source_kind":"arxiv","source_id":"2607.27283","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":"2026-07-31T00:10:29Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"RGVdjqPzBZo5NJmdDHJX20P3nBhS9G4z0J35TxO+/10X+mCAePsw8u+S8YCEQGDPCZyauNy9HZ2uOatQNg4xCQ==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-08-02T18:03:18.160778Z"},"content_sha256":"a1dc64af931aa6cb19b9787ff846d8b112ae9fc27ee4f466413490d3acd5ed22","schema_version":"1.0","event_id":"sha256:a1dc64af931aa6cb19b9787ff846d8b112ae9fc27ee4f466413490d3acd5ed22"},{"event_type":"graph_snapshot","subject_pith_number":"pith:2026:WNSH2SFNBIHTIOCFFSFHULXAZQ","target":"graph","payload":{"graph_snapshot":{"paper":{"title":"Benchmarking the Residual: What Long-Horizon Evaluations Add Beyond Matched Short-Task Performance","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.AI","cs.SE"],"primary_cat":"cs.LG","authors_text":"Chao Peng, Hande Dong, Peijie Dong, Qiang Lin, Zhiheng Lyu","submitted_at":"2026-07-29T13:57:32Z","abstract_excerpt":"Long-horizon benchmarks often show that agents fail more as tasks become longer. This observation is useful for deployment, but it does not by itself explain why failure occurs. More stages create more opportunities for ordinary errors to compound; longer tasks may also contain harder individual decisions or become harder as conversation history, tool outputs, and environment changes accumulate. We use trajectory-induced degradation to mean this last possibility: earlier execution makes later work harder. When the harmful accumulation is specifically the text visible to the model, it is often "},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2607.27283","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/2607.27283/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":"2026-07-31T00:10:29Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"lpL4yf5l6wmqViczqSbMfgh1y672H9TQq5kD3PmXj6KZBeKM59YP5PZaw/T1NAU6WEM8TwR2zkR0t0vshgZMBA==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-08-02T18:03:18.161252Z"},"content_sha256":"079dc534d0c0da74fca4daaa9e4671a5321f1577e30afd5e02e5a137ad81c035","schema_version":"1.0","event_id":"sha256:079dc534d0c0da74fca4daaa9e4671a5321f1577e30afd5e02e5a137ad81c035"}],"timestamp_proofs":[],"mirror_hints":[{"mirror_type":"https","name":"Pith Resolver","base_url":"https://pith.science","bundle_url":"https://pith.science/pith/WNSH2SFNBIHTIOCFFSFHULXAZQ/bundle.json","state_url":"https://pith.science/pith/WNSH2SFNBIHTIOCFFSFHULXAZQ/state.json","well_known_bundle_url":"https://pith.science/.well-known/pith/WNSH2SFNBIHTIOCFFSFHULXAZQ/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":"2026-08-02T18:03:18Z","links":{"resolver":"https://pith.science/pith/WNSH2SFNBIHTIOCFFSFHULXAZQ","bundle":"https://pith.science/pith/WNSH2SFNBIHTIOCFFSFHULXAZQ/bundle.json","state":"https://pith.science/pith/WNSH2SFNBIHTIOCFFSFHULXAZQ/state.json","well_known_bundle":"https://pith.science/.well-known/pith/WNSH2SFNBIHTIOCFFSFHULXAZQ/bundle.json"},"state":{"state_type":"pith_open_graph_state","state_version":"1.0","pith_number":"pith:2026:WNSH2SFNBIHTIOCFFSFHULXAZQ","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":"3c8dad46a4fff4e8c3b267cf261b5f46815da42d157decbd67802bd6c14d1ca1","cross_cats_sorted":["cs.AI","cs.SE"],"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.LG","submitted_at":"2026-07-29T13:57:32Z","title_canon_sha256":"7ef9f2c8fcd9a29097520cf3d7289d4215c7772b64bceead9eb55a677035602b"},"schema_version":"1.0","source":{"id":"2607.27283","kind":"arxiv","version":1}},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2607.27283","created_at":"2026-07-31T00:10:29Z"},{"alias_kind":"arxiv_version","alias_value":"2607.27283v1","created_at":"2026-07-31T00:10:29Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2607.27283","created_at":"2026-07-31T00:10:29Z"},{"alias_kind":"pith_short_12","alias_value":"WNSH2SFNBIHT","created_at":"2026-07-31T00:10:29Z"},{"alias_kind":"pith_short_16","alias_value":"WNSH2SFNBIHTIOCF","created_at":"2026-07-31T00:10:29Z"},{"alias_kind":"pith_short_8","alias_value":"WNSH2SFN","created_at":"2026-07-31T00:10:29Z"}],"graph_snapshots":[{"event_id":"sha256:079dc534d0c0da74fca4daaa9e4671a5321f1577e30afd5e02e5a137ad81c035","target":"graph","created_at":"2026-07-31T00:10:29Z","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/2607.27283/integrity.json","findings":[],"snapshot_sha256":"c28c3603d3b5d939e8dc4c7e95fa8dfce3d595e45f758748cecf8e644a296938","summary":{"advisory":0,"by_detector":{},"critical":0,"informational":0}},"paper":{"abstract_excerpt":"Long-horizon benchmarks often show that agents fail more as tasks become longer. This observation is useful for deployment, but it does not by itself explain why failure occurs. More stages create more opportunities for ordinary errors to compound; longer tasks may also contain harder individual decisions or become harder as conversation history, tool outputs, and environment changes accumulate. We use trajectory-induced degradation to mean this last possibility: earlier execution makes later work harder. When the harmful accumulation is specifically the text visible to the model, it is often ","authors_text":"Chao Peng, Hande Dong, Peijie Dong, Qiang Lin, Zhiheng Lyu","cross_cats":["cs.AI","cs.SE"],"headline":"","license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.LG","submitted_at":"2026-07-29T13:57:32Z","title":"Benchmarking the Residual: What Long-Horizon Evaluations Add Beyond Matched Short-Task Performance"},"references":{"count":0,"internal_anchors":0,"resolved_work":0,"sample":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2607.27283","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:a1dc64af931aa6cb19b9787ff846d8b112ae9fc27ee4f466413490d3acd5ed22","target":"record","created_at":"2026-07-31T00:10:29Z","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":"3c8dad46a4fff4e8c3b267cf261b5f46815da42d157decbd67802bd6c14d1ca1","cross_cats_sorted":["cs.AI","cs.SE"],"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.LG","submitted_at":"2026-07-29T13:57:32Z","title_canon_sha256":"7ef9f2c8fcd9a29097520cf3d7289d4215c7772b64bceead9eb55a677035602b"},"schema_version":"1.0","source":{"id":"2607.27283","kind":"arxiv","version":1}},"canonical_sha256":"b3647d48ad0a0f3438452c8a7a2ee0cc2914f610f48cfbf1b6bb2dcf162418d4","receipt":{"builder_version":"pith-number-builder-2026-05-17-v1","canonical_sha256":"b3647d48ad0a0f3438452c8a7a2ee0cc2914f610f48cfbf1b6bb2dcf162418d4","first_computed_at":"2026-07-31T00:10:29.953952Z","kind":"pith_receipt","last_reissued_at":"2026-07-31T00:10:29.953952Z","receipt_version":"0.3","signature_status":"unsigned_v0"},"source_id":"2607.27283","source_kind":"arxiv","source_version":1}}},"equivocations":[],"invalid_events":[],"applied_event_ids":["sha256:a1dc64af931aa6cb19b9787ff846d8b112ae9fc27ee4f466413490d3acd5ed22","sha256:079dc534d0c0da74fca4daaa9e4671a5321f1577e30afd5e02e5a137ad81c035"],"state_sha256":"45e1657cb1e8425eba0057aae7975cbca8a7d9b180ebe25139640efd8957137c"},"bundle_signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"tBf4tOo3l0BehmyETNUbX8s7/2Nm/p5mFMTvWSgJ4Wsy+FFDzEcBV9H4gKqkxrV7zUWv6OU7VQFG+tifKPdHBg==","signed_message":"bundle_sha256_bytes","signed_at":"2026-08-02T18:03:18.164464Z","bundle_sha256":"fe9dd682c75bc2c2bd04dc2953502b19ff746aa0c9ceb016ceb1544dc9273e0a"}}