{"state_type":"pith_open_graph_state","state_version":"1.0","pith_number":"pith:2025:JQK5FKKVWUSAYQJYIHHMMZAL7V","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":"418b96fc7d06a153bdf1ccc6cdd918007299bdbcf8752b39fd51bec47c82dd51","cross_cats_sorted":["cs.AI"],"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CL","submitted_at":"2025-06-06T17:52:18Z","title_canon_sha256":"98e3b0a7e3a7fe0f5d3e00770b7a4c31721100fce1b4df4f7f94195824854ce7"},"schema_version":"1.0","source":{"id":"2506.11102","kind":"arxiv","version":1}},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2506.11102","created_at":"2026-07-05T11:20:59Z"},{"alias_kind":"arxiv_version","alias_value":"2506.11102v1","created_at":"2026-07-05T11:20:59Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2506.11102","created_at":"2026-07-05T11:20:59Z"},{"alias_kind":"pith_short_12","alias_value":"JQK5FKKVWUSA","created_at":"2026-07-05T11:20:59Z"},{"alias_kind":"pith_short_16","alias_value":"JQK5FKKVWUSAYQJY","created_at":"2026-07-05T11:20:59Z"},{"alias_kind":"pith_short_8","alias_value":"JQK5FKKV","created_at":"2026-07-05T11:20:59Z"}],"graph_snapshots":[{"event_id":"sha256:2452345786c28f0911df190aeb90fc5cd5b0202b6a3e8cfafa5688f1aebc410f","target":"graph","created_at":"2026-07-05T11:20:59Z","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/2506.11102/integrity.json","findings":[],"snapshot_sha256":"c28c3603d3b5d939e8dc4c7e95fa8dfce3d595e45f758748cecf8e644a296938","summary":{"advisory":0,"by_detector":{},"critical":0,"informational":0}},"paper":{"abstract_excerpt":"The advent of large language models (LLMs), such as GPT, Gemini, and DeepSeek, has significantly advanced natural language processing, giving rise to sophisticated chatbots capable of diverse language-related tasks. The transition from these traditional LLM chatbots to more advanced AI agents represents a pivotal evolutionary step. However, existing evaluation frameworks often blur the distinctions between LLM chatbots and AI agents, leading to confusion among researchers selecting appropriate benchmarks. To bridge this gap, this paper introduces a systematic analysis of current evaluation app","authors_text":"Bo Chen, Congmin Zheng, Jiachen Zhu, Jianghao Lin, Menghui Zhu, Renting Rui, Rong Shan, Ruiming Tang, Weinan Zhang, Weiwen Liu, Yong Yu, Yunjia Xi","cross_cats":["cs.AI"],"headline":"","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CL","submitted_at":"2025-06-06T17:52:18Z","title":"Evolutionary Perspectives on the Evaluation of LLM-Based AI Agents: A Comprehensive Survey"},"references":{"count":0,"internal_anchors":0,"resolved_work":0,"sample":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2506.11102","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":"sha256:325a65784841a02f6232ca8a20d7c5f794654843abd09f04a908aaae4673e0ef","target":"record","created_at":"2026-07-05T11:20:59Z","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":"418b96fc7d06a153bdf1ccc6cdd918007299bdbcf8752b39fd51bec47c82dd51","cross_cats_sorted":["cs.AI"],"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CL","submitted_at":"2025-06-06T17:52:18Z","title_canon_sha256":"98e3b0a7e3a7fe0f5d3e00770b7a4c31721100fce1b4df4f7f94195824854ce7"},"schema_version":"1.0","source":{"id":"2506.11102","kind":"arxiv","version":1}},"canonical_sha256":"4c15d2a955b5240c413841cec6640bfd5fde57b6b27972d2d01e44eb041c42e6","receipt":{"algorithm":"ed25519","builder_version":"pith-number-builder-2026-05-17-v1","canonical_sha256":"4c15d2a955b5240c413841cec6640bfd5fde57b6b27972d2d01e44eb041c42e6","first_computed_at":"2026-07-05T11:20:59.475772Z","key_id":"pith-v1-2026-05","kind":"pith_receipt","last_reissued_at":"2026-07-05T11:20:59.475772Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","receipt_version":"0.3","signature_b64":"QDC4UyNDR1T/+MgGWxAqDj3Ex9QxCOLt9ucU8kgZN7r62K/X5gVV/7wchi4SmgHR1WNN1cT+QPFAUD8EVuSzAg==","signature_status":"signed_v1","signed_at":"2026-07-05T11:20:59.476291Z","signed_message":"canonical_sha256_bytes"},"source_id":"2506.11102","source_kind":"arxiv","source_version":1}}},"equivocations":[],"invalid_events":[],"applied_event_ids":["sha256:325a65784841a02f6232ca8a20d7c5f794654843abd09f04a908aaae4673e0ef","sha256:2452345786c28f0911df190aeb90fc5cd5b0202b6a3e8cfafa5688f1aebc410f"],"state_sha256":"99cff453339d18a7b2463c70b51206ca79e3c2b7424f1f4ccc4ea0adbf89f157"}