{"state_type":"pith_open_graph_state","state_version":"1.0","pith_number":"pith:2026:SDVJB5LAQSE3TNSM6XBLDSL7NJ","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":"cbaac8bf5b2c4d912f46099696b00364e33fc32a0facad5f87b181f978df4179","cross_cats_sorted":[],"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CL","submitted_at":"2026-01-15T15:28:21Z","title_canon_sha256":"3d979b6e5d362d05fa775d396840897d9e64cf49fa731b3320d7885a14b8bddf"},"schema_version":"1.0","source":{"id":"2601.10504","kind":"arxiv","version":2}},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2601.10504","created_at":"2026-07-10T01:18:57Z"},{"alias_kind":"arxiv_version","alias_value":"2601.10504v2","created_at":"2026-07-10T01:18:57Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2601.10504","created_at":"2026-07-10T01:18:57Z"},{"alias_kind":"pith_short_12","alias_value":"SDVJB5LAQSE3","created_at":"2026-07-10T01:18:57Z"},{"alias_kind":"pith_short_16","alias_value":"SDVJB5LAQSE3TNSM","created_at":"2026-07-10T01:18:57Z"},{"alias_kind":"pith_short_8","alias_value":"SDVJB5LA","created_at":"2026-07-10T01:18:57Z"}],"graph_snapshots":[{"event_id":"sha256:143ae12e1e9b8780885ac973c9c94ab3ee54683670864bbca2d683103ff9cef9","target":"graph","created_at":"2026-07-10T01:18:57Z","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/2601.10504/integrity.json","findings":[],"snapshot_sha256":"c28c3603d3b5d939e8dc4c7e95fa8dfce3d595e45f758748cecf8e644a296938","summary":{"advisory":0,"by_detector":{},"critical":0,"informational":0}},"paper":{"abstract_excerpt":"As Large Language Models (LLMs) increasingly operate as Deep Research (DR) Agents capable of autonomous investigation and information synthesis, reliable evaluation of their task performance has become a critical bottleneck. Current benchmarks predominantly rely on static datasets, which suffer from several limitations: limited task generality, temporal misalignment, and data contamination. To address these, we introduce DR-Arena, a fully automated evaluation framework that pushes DR agents to their capability limits through dynamic investigation. DR-Arena constructs real-time Information Tree","authors_text":"Ruochen Zhao, Wenxuan Zhang, Yang Deng, Yiwen Gao","cross_cats":[],"headline":"","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CL","submitted_at":"2026-01-15T15:28:21Z","title":"DR-Arena: an Automated Evaluation Framework for Deep Research Agents"},"references":{"count":0,"internal_anchors":0,"resolved_work":0,"sample":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2601.10504","kind":"arxiv","version":2},"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":"sha256:3672ee8695b67150243b09f178c23c53b6ca8e57d955a6f7c8aa13f17c56508d","target":"record","created_at":"2026-07-10T01:18:57Z","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":"cbaac8bf5b2c4d912f46099696b00364e33fc32a0facad5f87b181f978df4179","cross_cats_sorted":[],"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CL","submitted_at":"2026-01-15T15:28:21Z","title_canon_sha256":"3d979b6e5d362d05fa775d396840897d9e64cf49fa731b3320d7885a14b8bddf"},"schema_version":"1.0","source":{"id":"2601.10504","kind":"arxiv","version":2}},"canonical_sha256":"90ea90f5608489b9b64cf5c2b1c97f6a40e1132ba96e8b7be9f6386bf2218502","receipt":{"algorithm":"ed25519","builder_version":"pith-number-builder-2026-05-17-v1","canonical_sha256":"90ea90f5608489b9b64cf5c2b1c97f6a40e1132ba96e8b7be9f6386bf2218502","first_computed_at":"2026-07-10T01:18:57.439137Z","key_id":"pith-v1-2026-05","kind":"pith_receipt","last_reissued_at":"2026-07-10T01:18:57.439137Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","receipt_version":"0.3","signature_b64":"IUj7i5la+iPv9duwRK1VYpOCyBGpzz6KY9UY8WWg63xYilpHE/BCIgWzTFi3LFO5N1nekNKErn7hczoS9l67Aw==","signature_status":"signed_v1","signed_at":"2026-07-10T01:18:57.439648Z","signed_message":"canonical_sha256_bytes"},"source_id":"2601.10504","source_kind":"arxiv","source_version":2}}},"equivocations":[],"invalid_events":[],"applied_event_ids":["sha256:3672ee8695b67150243b09f178c23c53b6ca8e57d955a6f7c8aa13f17c56508d","sha256:143ae12e1e9b8780885ac973c9c94ab3ee54683670864bbca2d683103ff9cef9"],"state_sha256":"7767f000a39953b229e1fbd080f818a7f2da154628c014c68a0d2af02e970cdb"}