{"bundle_type":"pith_open_graph_bundle","bundle_version":"1.0","pith_number":"pith:2026:77NRCAIXBNTTJWLUDRZHIRZHFD","short_pith_number":"pith:77NRCAIX","canonical_record":{"source":{"id":"2606.12736","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/licenses/by-nc-sa/4.0/","primary_cat":"cs.AI","submitted_at":"2026-06-10T22:55:30Z","cross_cats_sorted":["cs.LG"],"title_canon_sha256":"1c1cbde117c891385e6375fb00469bce3dbab518987adad66bf76791ae304c4c","abstract_canon_sha256":"4f8df36ff8369d2ebe4c684d0529cd28f7ffd04c74f82b347b8671ef309fe3b9"},"schema_version":"1.0"},"canonical_sha256":"ffdb1101170b6734d9741c7274472728c2a5c5fe5002f75d21ff9fd3897ceb1a","source":{"kind":"arxiv","id":"2606.12736","version":1},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2606.12736","created_at":"2026-06-12T01:08:48Z"},{"alias_kind":"arxiv_version","alias_value":"2606.12736v1","created_at":"2026-06-12T01:08:48Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2606.12736","created_at":"2026-06-12T01:08:48Z"},{"alias_kind":"pith_short_12","alias_value":"77NRCAIXBNTT","created_at":"2026-06-12T01:08:48Z"},{"alias_kind":"pith_short_16","alias_value":"77NRCAIXBNTTJWLU","created_at":"2026-06-12T01:08:48Z"},{"alias_kind":"pith_short_8","alias_value":"77NRCAIX","created_at":"2026-06-12T01:08:48Z"}],"events":[{"event_type":"record_created","subject_pith_number":"pith:2026:77NRCAIXBNTTJWLUDRZHIRZHFD","target":"record","payload":{"canonical_record":{"source":{"id":"2606.12736","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/licenses/by-nc-sa/4.0/","primary_cat":"cs.AI","submitted_at":"2026-06-10T22:55:30Z","cross_cats_sorted":["cs.LG"],"title_canon_sha256":"1c1cbde117c891385e6375fb00469bce3dbab518987adad66bf76791ae304c4c","abstract_canon_sha256":"4f8df36ff8369d2ebe4c684d0529cd28f7ffd04c74f82b347b8671ef309fe3b9"},"schema_version":"1.0"},"canonical_sha256":"ffdb1101170b6734d9741c7274472728c2a5c5fe5002f75d21ff9fd3897ceb1a","receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-06-12T01:08:48.418433Z","signature_b64":"1WOVM1VDyiF+eKFOCiSmla3iQlv1aP4HZ7MUSNZsv8LZa+xl+SeUYJ714kWt1iZ3ycTMt9qbmu3JMPSZUB9oAQ==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"ffdb1101170b6734d9741c7274472728c2a5c5fe5002f75d21ff9fd3897ceb1a","last_reissued_at":"2026-06-12T01:08:48.417466Z","signature_status":"signed_v1","first_computed_at":"2026-06-12T01:08:48.417466Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"source_kind":"arxiv","source_id":"2606.12736","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":"2026-06-12T01:08:48Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"1CtCo4LzTyGoqUiQVGOoYwVgooeyfz2DYWEtbEmf2Feiw6D5O6kWznWgUoFcZ13g/Zv4VxJasbCuBZrEN2W6Dw==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-06-27T18:57:14.471831Z"},"content_sha256":"3b634907d3f7cacb8d7e987fcb15d933b01433fce978c35a63b4abe06c1c2581","schema_version":"1.0","event_id":"sha256:3b634907d3f7cacb8d7e987fcb15d933b01433fce978c35a63b4abe06c1c2581"},{"event_type":"graph_snapshot","subject_pith_number":"pith:2026:77NRCAIXBNTTJWLUDRZHIRZHFD","target":"graph","payload":{"graph_snapshot":{"paper":{"title":"Benchmarking AI Agents for Addressing Scientific Challenges Across Scales","license":"http://creativecommons.org/licenses/by-nc-sa/4.0/","headline":"","cross_cats":["cs.LG"],"primary_cat":"cs.AI","authors_text":"Ada Fang, Allen Xin Wang, Antonia Panescu, Arman Cohan, Botao Yu, Haoran Shao, Hongyu Zhao, Hua Xu, James Zou, Jihang Chen, Kaize Ding, Kunyang Sun, Leqi Xu, Lingzhou Xue, Lisa Xinyi Chen, Marinka Zitnik, Qingyu Chen, Rex Ying, Sihan Jiang, Siyi Gu, Siyu Chen, Tianyu Liu, Wangjie Zheng, Wengong Jin, Wenxin Long, Xinyang Hu, Xinyu Wei, Yuanqi Du, Yueqian Jing, Zhiyuan Cao, Zhuoran Yang, Ziqing Wang, Ziyao Zeng","submitted_at":"2026-06-10T22:55:30Z","abstract_excerpt":"AI agents are increasingly being developed to accelerate scientific discovery, yet their practical capabilities in real research settings remain poorly understood. Existing benchmarks for AI agents rarely capture the complexity, heterogeneity, and extended reasoning required by scientific work, whereas benchmarks for scientific tasks often reduce research to static, direct problems and provide limited support for interactive evaluation. Here, we introduce SciAgentArena, a systematic benchmark for evaluating AI agents in real-world scientific research scenarios drawn from emerging needs across "},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2606.12736","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/2606.12736/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":"2026-06-12T01:08:48Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"iCS2J5HlqQZ5sbdZPq0nq3o+U0ig/iwZJtYOfviFGE3n1LD4J0nbrzOL0g/kri+kA6ZDZrCoXoY5oBkOrWSPAg==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-06-27T18:57:14.472210Z"},"content_sha256":"e1120fa8257b84e437466174ff30e534f0eca63aeeafa49c8dcb3dff253dad77","schema_version":"1.0","event_id":"sha256:e1120fa8257b84e437466174ff30e534f0eca63aeeafa49c8dcb3dff253dad77"}],"timestamp_proofs":[],"mirror_hints":[{"mirror_type":"https","name":"Pith Resolver","base_url":"https://pith.science","bundle_url":"https://pith.science/pith/77NRCAIXBNTTJWLUDRZHIRZHFD/bundle.json","state_url":"https://pith.science/pith/77NRCAIXBNTTJWLUDRZHIRZHFD/state.json","well_known_bundle_url":"https://pith.science/.well-known/pith/77NRCAIXBNTTJWLUDRZHIRZHFD/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":"2026-06-27T18:57:14Z","links":{"resolver":"https://pith.science/pith/77NRCAIXBNTTJWLUDRZHIRZHFD","bundle":"https://pith.science/pith/77NRCAIXBNTTJWLUDRZHIRZHFD/bundle.json","state":"https://pith.science/pith/77NRCAIXBNTTJWLUDRZHIRZHFD/state.json","well_known_bundle":"https://pith.science/.well-known/pith/77NRCAIXBNTTJWLUDRZHIRZHFD/bundle.json"},"state":{"state_type":"pith_open_graph_state","state_version":"1.0","pith_number":"pith:2026:77NRCAIXBNTTJWLUDRZHIRZHFD","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":"4f8df36ff8369d2ebe4c684d0529cd28f7ffd04c74f82b347b8671ef309fe3b9","cross_cats_sorted":["cs.LG"],"license":"http://creativecommons.org/licenses/by-nc-sa/4.0/","primary_cat":"cs.AI","submitted_at":"2026-06-10T22:55:30Z","title_canon_sha256":"1c1cbde117c891385e6375fb00469bce3dbab518987adad66bf76791ae304c4c"},"schema_version":"1.0","source":{"id":"2606.12736","kind":"arxiv","version":1}},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2606.12736","created_at":"2026-06-12T01:08:48Z"},{"alias_kind":"arxiv_version","alias_value":"2606.12736v1","created_at":"2026-06-12T01:08:48Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2606.12736","created_at":"2026-06-12T01:08:48Z"},{"alias_kind":"pith_short_12","alias_value":"77NRCAIXBNTT","created_at":"2026-06-12T01:08:48Z"},{"alias_kind":"pith_short_16","alias_value":"77NRCAIXBNTTJWLU","created_at":"2026-06-12T01:08:48Z"},{"alias_kind":"pith_short_8","alias_value":"77NRCAIX","created_at":"2026-06-12T01:08:48Z"}],"graph_snapshots":[{"event_id":"sha256:e1120fa8257b84e437466174ff30e534f0eca63aeeafa49c8dcb3dff253dad77","target":"graph","created_at":"2026-06-12T01:08:48Z","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/2606.12736/integrity.json","findings":[],"snapshot_sha256":"c28c3603d3b5d939e8dc4c7e95fa8dfce3d595e45f758748cecf8e644a296938","summary":{"advisory":0,"by_detector":{},"critical":0,"informational":0}},"paper":{"abstract_excerpt":"AI agents are increasingly being developed to accelerate scientific discovery, yet their practical capabilities in real research settings remain poorly understood. Existing benchmarks for AI agents rarely capture the complexity, heterogeneity, and extended reasoning required by scientific work, whereas benchmarks for scientific tasks often reduce research to static, direct problems and provide limited support for interactive evaluation. Here, we introduce SciAgentArena, a systematic benchmark for evaluating AI agents in real-world scientific research scenarios drawn from emerging needs across ","authors_text":"Ada Fang, Allen Xin Wang, Antonia Panescu, Arman Cohan, Botao Yu, Haoran Shao, Hongyu Zhao, Hua Xu, James Zou, Jihang Chen, Kaize Ding, Kunyang Sun, Leqi Xu, Lingzhou Xue, Lisa Xinyi Chen, Marinka Zitnik, Qingyu Chen, Rex Ying, Sihan Jiang, Siyi Gu, Siyu Chen, Tianyu Liu, Wangjie Zheng, Wengong Jin, Wenxin Long, Xinyang Hu, Xinyu Wei, Yuanqi Du, Yueqian Jing, Zhiyuan Cao, Zhuoran Yang, Ziqing Wang, Ziyao Zeng","cross_cats":["cs.LG"],"headline":"","license":"http://creativecommons.org/licenses/by-nc-sa/4.0/","primary_cat":"cs.AI","submitted_at":"2026-06-10T22:55:30Z","title":"Benchmarking AI Agents for Addressing Scientific Challenges Across Scales"},"references":{"count":0,"internal_anchors":0,"resolved_work":0,"sample":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2606.12736","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":"sha256:3b634907d3f7cacb8d7e987fcb15d933b01433fce978c35a63b4abe06c1c2581","target":"record","created_at":"2026-06-12T01:08:48Z","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":"4f8df36ff8369d2ebe4c684d0529cd28f7ffd04c74f82b347b8671ef309fe3b9","cross_cats_sorted":["cs.LG"],"license":"http://creativecommons.org/licenses/by-nc-sa/4.0/","primary_cat":"cs.AI","submitted_at":"2026-06-10T22:55:30Z","title_canon_sha256":"1c1cbde117c891385e6375fb00469bce3dbab518987adad66bf76791ae304c4c"},"schema_version":"1.0","source":{"id":"2606.12736","kind":"arxiv","version":1}},"canonical_sha256":"ffdb1101170b6734d9741c7274472728c2a5c5fe5002f75d21ff9fd3897ceb1a","receipt":{"algorithm":"ed25519","builder_version":"pith-number-builder-2026-05-17-v1","canonical_sha256":"ffdb1101170b6734d9741c7274472728c2a5c5fe5002f75d21ff9fd3897ceb1a","first_computed_at":"2026-06-12T01:08:48.417466Z","key_id":"pith-v1-2026-05","kind":"pith_receipt","last_reissued_at":"2026-06-12T01:08:48.417466Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","receipt_version":"0.3","signature_b64":"1WOVM1VDyiF+eKFOCiSmla3iQlv1aP4HZ7MUSNZsv8LZa+xl+SeUYJ714kWt1iZ3ycTMt9qbmu3JMPSZUB9oAQ==","signature_status":"signed_v1","signed_at":"2026-06-12T01:08:48.418433Z","signed_message":"canonical_sha256_bytes"},"source_id":"2606.12736","source_kind":"arxiv","source_version":1}}},"equivocations":[],"invalid_events":[],"applied_event_ids":["sha256:3b634907d3f7cacb8d7e987fcb15d933b01433fce978c35a63b4abe06c1c2581","sha256:e1120fa8257b84e437466174ff30e534f0eca63aeeafa49c8dcb3dff253dad77"],"state_sha256":"63cc6de9cf93bbdb20a3c28440b44c07c00237500c9fa97947e3f33eb6128f1d"},"bundle_signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"RfaE7NK/Va0BHj3gNLU//hTlJOf5nYlDB0ROsSVDzz0zG0/YhRG24aMxWQ0grEzHExJBx5V4kT8bt2Lp3PxwBg==","signed_message":"bundle_sha256_bytes","signed_at":"2026-06-27T18:57:14.474136Z","bundle_sha256":"08e45ffead49e3a11b58cc5a73c90754aa776f6ad58d4e992177dc7c10d88c27"}}