{"state_type":"pith_open_graph_state","state_version":"1.0","pith_number":"pith:2025:5O3KPRVQSD3HV2UYXUOVN65LJN","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":"54f54c3bf1ae7bdb9274bcf23b4002fbd0146075c80f73b1fc57b122d0de022a","cross_cats_sorted":[],"license":"http://creativecommons.org/licenses/by-sa/4.0/","primary_cat":"cs.CL","submitted_at":"2025-04-10T20:39:18Z","title_canon_sha256":"c649e284bf4679c16522599551a098c99ee9b685e2068ea230d212561816b22d"},"schema_version":"1.0","source":{"id":"2504.08120","kind":"arxiv","version":3}},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2504.08120","created_at":"2026-07-05T11:12:30Z"},{"alias_kind":"arxiv_version","alias_value":"2504.08120v3","created_at":"2026-07-05T11:12:30Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2504.08120","created_at":"2026-07-05T11:12:30Z"},{"alias_kind":"pith_short_12","alias_value":"5O3KPRVQSD3H","created_at":"2026-07-05T11:12:30Z"},{"alias_kind":"pith_short_16","alias_value":"5O3KPRVQSD3HV2UY","created_at":"2026-07-05T11:12:30Z"},{"alias_kind":"pith_short_8","alias_value":"5O3KPRVQ","created_at":"2026-07-05T11:12:30Z"}],"graph_snapshots":[{"event_id":"sha256:71fc9a90f3444c1d22a6581f8100a112837785d81b900cf5bce6157039e23049","target":"graph","created_at":"2026-07-05T11:12:30Z","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/2504.08120/integrity.json","findings":[],"snapshot_sha256":"c28c3603d3b5d939e8dc4c7e95fa8dfce3d595e45f758748cecf8e644a296938","summary":{"advisory":0,"by_detector":{},"critical":0,"informational":0}},"paper":{"abstract_excerpt":"Reasoning-enabled large language models (LLMs) excel in logical tasks, yet their utility for evaluating natural language generation remains unexplored. This study systematically compares reasoning LLMs with non-reasoning counterparts across machine translation and text summarization evaluation tasks. We evaluate eight models spanning state-of-the-art reasoning models (DeepSeek-R1, OpenAI o3), their distilled variants (8B-70B parameters), and equivalent non-reasoning LLMs. Experiments on WMT23 and SummEval benchmarks reveal architecture and task-dependent benefits: OpenAI o3-mini models show im","authors_text":"Christian Greisinger, Christoph Leiter, Daniil Larionov, Ran Zhang, Sotaro Takeshita, Steffen Eger, Yanran Chen, Zhipin Wang","cross_cats":[],"headline":"","license":"http://creativecommons.org/licenses/by-sa/4.0/","primary_cat":"cs.CL","submitted_at":"2025-04-10T20:39:18Z","title":"DeepSeek-R1 vs. o3-mini: How Well can Reasoning LLMs Evaluate MT and Summarization?"},"references":{"count":0,"internal_anchors":0,"resolved_work":0,"sample":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2504.08120","kind":"arxiv","version":3},"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":"sha256:2f12089a0f83b3b220b93a2b49edb6d17a75eccc5f84dd49b73ad9f99383bdfd","target":"record","created_at":"2026-07-05T11:12:30Z","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":"54f54c3bf1ae7bdb9274bcf23b4002fbd0146075c80f73b1fc57b122d0de022a","cross_cats_sorted":[],"license":"http://creativecommons.org/licenses/by-sa/4.0/","primary_cat":"cs.CL","submitted_at":"2025-04-10T20:39:18Z","title_canon_sha256":"c649e284bf4679c16522599551a098c99ee9b685e2068ea230d212561816b22d"},"schema_version":"1.0","source":{"id":"2504.08120","kind":"arxiv","version":3}},"canonical_sha256":"ebb6a7c6b090f67aea98bd1d56fbab4b470fd73abee2a4da3e60587dc245162b","receipt":{"algorithm":"ed25519","builder_version":"pith-number-builder-2026-05-17-v1","canonical_sha256":"ebb6a7c6b090f67aea98bd1d56fbab4b470fd73abee2a4da3e60587dc245162b","first_computed_at":"2026-07-05T11:12:30.628723Z","key_id":"pith-v1-2026-05","kind":"pith_receipt","last_reissued_at":"2026-07-05T11:12:30.628723Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","receipt_version":"0.3","signature_b64":"3YA6vHEAjHCeJ9BBjbSCkCXfrtheyDusauX3miGHyDCWztE+plfkKRrKlJ0UETLh/pMn9yZQSFnWlc6uz8U1BA==","signature_status":"signed_v1","signed_at":"2026-07-05T11:12:30.629259Z","signed_message":"canonical_sha256_bytes"},"source_id":"2504.08120","source_kind":"arxiv","source_version":3}}},"equivocations":[],"invalid_events":[],"applied_event_ids":["sha256:2f12089a0f83b3b220b93a2b49edb6d17a75eccc5f84dd49b73ad9f99383bdfd","sha256:71fc9a90f3444c1d22a6581f8100a112837785d81b900cf5bce6157039e23049"],"state_sha256":"5bd090926c0b3a14e520fdb17306632fe1201c5eba391a91817562489d660453"}