{"bundle_type":"pith_open_graph_bundle","bundle_version":"1.0","pith_number":"pith:2025:FBZOOOK6RAEYJII2XSAL5KICQN","short_pith_number":"pith:FBZOOOK6","canonical_record":{"source":{"id":"2505.22661","kind":"arxiv","version":1},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CL","submitted_at":"2025-05-28T17:59:43Z","cross_cats_sorted":[],"title_canon_sha256":"2fa923496e93b604874664529db62a5d75bee3348bdb40c1681d56adca2d6560","abstract_canon_sha256":"d8136084124fb976d7ec64b11f3dd1ad4539d892e7eb329e3265d83c3460cbcf"},"schema_version":"1.0"},"canonical_sha256":"2872e7395e880984a11abc80bea90283617d1140dc3f85c88b0230a115f26e19","source":{"kind":"arxiv","id":"2505.22661","version":1},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2505.22661","created_at":"2026-07-05T11:11:28Z"},{"alias_kind":"arxiv_version","alias_value":"2505.22661v1","created_at":"2026-07-05T11:11:28Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2505.22661","created_at":"2026-07-05T11:11:28Z"},{"alias_kind":"pith_short_12","alias_value":"FBZOOOK6RAEY","created_at":"2026-07-05T11:11:28Z"},{"alias_kind":"pith_short_16","alias_value":"FBZOOOK6RAEYJII2","created_at":"2026-07-05T11:11:28Z"},{"alias_kind":"pith_short_8","alias_value":"FBZOOOK6","created_at":"2026-07-05T11:11:28Z"}],"events":[{"event_type":"record_created","subject_pith_number":"pith:2025:FBZOOOK6RAEYJII2XSAL5KICQN","target":"record","payload":{"canonical_record":{"source":{"id":"2505.22661","kind":"arxiv","version":1},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CL","submitted_at":"2025-05-28T17:59:43Z","cross_cats_sorted":[],"title_canon_sha256":"2fa923496e93b604874664529db62a5d75bee3348bdb40c1681d56adca2d6560","abstract_canon_sha256":"d8136084124fb976d7ec64b11f3dd1ad4539d892e7eb329e3265d83c3460cbcf"},"schema_version":"1.0"},"canonical_sha256":"2872e7395e880984a11abc80bea90283617d1140dc3f85c88b0230a115f26e19","receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T11:11:28.560110Z","signature_b64":"rMvOKdGjeyxnPedj7h14JbVGihLUNDh/2NBGxj1PYi5ARfqHhPAqLiBFs+zilSSdAuMckTTzGGRIvIN7SttGDw==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"2872e7395e880984a11abc80bea90283617d1140dc3f85c88b0230a115f26e19","last_reissued_at":"2026-07-05T11:11:28.559572Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T11:11:28.559572Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"source_kind":"arxiv","source_id":"2505.22661","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":"2026-07-05T11:11:28Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"GmBzINhIxS010meiKqhN56maqzn8Ov4if2mGnE3cQH+ll0FDr/BQtckvXo/I9nJ6awFE4x8emzhFaNB6dl3uCg==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-08-07T21:27:49.085275Z"},"content_sha256":"b1335648b34ed9d2900cbe1e94e2ca1d67cd970130558cc0acad8406fdc5036a","schema_version":"1.0","event_id":"sha256:b1335648b34ed9d2900cbe1e94e2ca1d67cd970130558cc0acad8406fdc5036a"},{"event_type":"graph_snapshot","subject_pith_number":"pith:2025:FBZOOOK6RAEYJII2XSAL5KICQN","target":"graph","payload":{"graph_snapshot":{"paper":{"title":"GuessArena: Guess Who I Am? A Self-Adaptive Framework for Evaluating LLMs in Domain-Specific Knowledge and Reasoning","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":[],"primary_cat":"cs.CL","authors_text":"Bo Tang, Ding Chen, Feiyu Xiong, Qingchen Yu, Simin Niu, Zhiyu Li, Zifan Zheng","submitted_at":"2025-05-28T17:59:43Z","abstract_excerpt":"The evaluation of large language models (LLMs) has traditionally relied on static benchmarks, a paradigm that poses two major limitations: (1) predefined test sets lack adaptability to diverse application domains, and (2) standardized evaluation protocols often fail to capture fine-grained assessments of domain-specific knowledge and contextual reasoning abilities. To overcome these challenges, we propose GuessArena, an adaptive evaluation framework grounded in adversarial game-based interactions. Inspired by the interactive structure of the Guess Who I Am? game, our framework seamlessly integ"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2505.22661","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/2505.22661/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":"2026-07-05T11:11:28Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"CWjoTDpFLhjy3pL7IsibeVUSpQrEO3GVFpcNN91QcNopGA52Ptnv+IvGEZn0hkkoNRVgG/I2uAaVgrlQLdiBAg==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-08-07T21:27:49.085774Z"},"content_sha256":"e92428db272a4ea472cfadffee286031d347e381d3ebf2279e1ead659e06dfa7","schema_version":"1.0","event_id":"sha256:e92428db272a4ea472cfadffee286031d347e381d3ebf2279e1ead659e06dfa7"}],"timestamp_proofs":[],"mirror_hints":[{"mirror_type":"https","name":"Pith Resolver","base_url":"https://pith.science","bundle_url":"https://pith.science/pith/FBZOOOK6RAEYJII2XSAL5KICQN/bundle.json","state_url":"https://pith.science/pith/FBZOOOK6RAEYJII2XSAL5KICQN/state.json","well_known_bundle_url":"https://pith.science/.well-known/pith/FBZOOOK6RAEYJII2XSAL5KICQN/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":"2026-08-07T21:27:49Z","links":{"resolver":"https://pith.science/pith/FBZOOOK6RAEYJII2XSAL5KICQN","bundle":"https://pith.science/pith/FBZOOOK6RAEYJII2XSAL5KICQN/bundle.json","state":"https://pith.science/pith/FBZOOOK6RAEYJII2XSAL5KICQN/state.json","well_known_bundle":"https://pith.science/.well-known/pith/FBZOOOK6RAEYJII2XSAL5KICQN/bundle.json"},"state":{"state_type":"pith_open_graph_state","state_version":"1.0","pith_number":"pith:2025:FBZOOOK6RAEYJII2XSAL5KICQN","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":"d8136084124fb976d7ec64b11f3dd1ad4539d892e7eb329e3265d83c3460cbcf","cross_cats_sorted":[],"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CL","submitted_at":"2025-05-28T17:59:43Z","title_canon_sha256":"2fa923496e93b604874664529db62a5d75bee3348bdb40c1681d56adca2d6560"},"schema_version":"1.0","source":{"id":"2505.22661","kind":"arxiv","version":1}},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2505.22661","created_at":"2026-07-05T11:11:28Z"},{"alias_kind":"arxiv_version","alias_value":"2505.22661v1","created_at":"2026-07-05T11:11:28Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2505.22661","created_at":"2026-07-05T11:11:28Z"},{"alias_kind":"pith_short_12","alias_value":"FBZOOOK6RAEY","created_at":"2026-07-05T11:11:28Z"},{"alias_kind":"pith_short_16","alias_value":"FBZOOOK6RAEYJII2","created_at":"2026-07-05T11:11:28Z"},{"alias_kind":"pith_short_8","alias_value":"FBZOOOK6","created_at":"2026-07-05T11:11:28Z"}],"graph_snapshots":[{"event_id":"sha256:e92428db272a4ea472cfadffee286031d347e381d3ebf2279e1ead659e06dfa7","target":"graph","created_at":"2026-07-05T11:11:28Z","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/2505.22661/integrity.json","findings":[],"snapshot_sha256":"c28c3603d3b5d939e8dc4c7e95fa8dfce3d595e45f758748cecf8e644a296938","summary":{"advisory":0,"by_detector":{},"critical":0,"informational":0}},"paper":{"abstract_excerpt":"The evaluation of large language models (LLMs) has traditionally relied on static benchmarks, a paradigm that poses two major limitations: (1) predefined test sets lack adaptability to diverse application domains, and (2) standardized evaluation protocols often fail to capture fine-grained assessments of domain-specific knowledge and contextual reasoning abilities. To overcome these challenges, we propose GuessArena, an adaptive evaluation framework grounded in adversarial game-based interactions. Inspired by the interactive structure of the Guess Who I Am? game, our framework seamlessly integ","authors_text":"Bo Tang, Ding Chen, Feiyu Xiong, Qingchen Yu, Simin Niu, Zhiyu Li, Zifan Zheng","cross_cats":[],"headline":"","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CL","submitted_at":"2025-05-28T17:59:43Z","title":"GuessArena: Guess Who I Am? A Self-Adaptive Framework for Evaluating LLMs in Domain-Specific Knowledge and Reasoning"},"references":{"count":0,"internal_anchors":0,"resolved_work":0,"sample":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2505.22661","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:b1335648b34ed9d2900cbe1e94e2ca1d67cd970130558cc0acad8406fdc5036a","target":"record","created_at":"2026-07-05T11:11:28Z","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":"d8136084124fb976d7ec64b11f3dd1ad4539d892e7eb329e3265d83c3460cbcf","cross_cats_sorted":[],"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CL","submitted_at":"2025-05-28T17:59:43Z","title_canon_sha256":"2fa923496e93b604874664529db62a5d75bee3348bdb40c1681d56adca2d6560"},"schema_version":"1.0","source":{"id":"2505.22661","kind":"arxiv","version":1}},"canonical_sha256":"2872e7395e880984a11abc80bea90283617d1140dc3f85c88b0230a115f26e19","receipt":{"algorithm":"ed25519","builder_version":"pith-number-builder-2026-05-17-v1","canonical_sha256":"2872e7395e880984a11abc80bea90283617d1140dc3f85c88b0230a115f26e19","first_computed_at":"2026-07-05T11:11:28.559572Z","key_id":"pith-v1-2026-05","kind":"pith_receipt","last_reissued_at":"2026-07-05T11:11:28.559572Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","receipt_version":"0.3","signature_b64":"rMvOKdGjeyxnPedj7h14JbVGihLUNDh/2NBGxj1PYi5ARfqHhPAqLiBFs+zilSSdAuMckTTzGGRIvIN7SttGDw==","signature_status":"signed_v1","signed_at":"2026-07-05T11:11:28.560110Z","signed_message":"canonical_sha256_bytes"},"source_id":"2505.22661","source_kind":"arxiv","source_version":1}}},"equivocations":[],"invalid_events":[],"applied_event_ids":["sha256:b1335648b34ed9d2900cbe1e94e2ca1d67cd970130558cc0acad8406fdc5036a","sha256:e92428db272a4ea472cfadffee286031d347e381d3ebf2279e1ead659e06dfa7"],"state_sha256":"7d11f8ba99daf84451bc3c19af6385aebf198a3b03c1b75036f33efe0f8ae368"},"bundle_signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"IwQ9zgGVws7IhPQX+qPi9esPG5Qqh2BAQd4L0kzGA02NGuLP5WdY2IQSRF+WTH76Tyu5ovPRx38y7qxO7NFOBg==","signed_message":"bundle_sha256_bytes","signed_at":"2026-08-07T21:27:49.090080Z","bundle_sha256":"68fdaafc162b70b32fed20d6d2670349c26cc3a0fc147a3e393371f23c7b171b"}}