{"bundle_type":"pith_open_graph_bundle","bundle_version":"1.0","pith_number":"pith:2025:JLAK3KLTEKWEO67Z55X5E7LBEJ","short_pith_number":"pith:JLAK3KLT","canonical_record":{"source":{"id":"2502.15815","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.LG","submitted_at":"2025-02-19T19:00:00Z","cross_cats_sorted":["astro-ph.CO","cs.AI","hep-ph","hep-th"],"title_canon_sha256":"5b93d434733e2ae2b5b4c9b6be2f928080ed23b1fd5d817fd6b594a8d4ed630c","abstract_canon_sha256":"956a41971638ae3a578f5da035c1ba690d2f85fd9befa7994af120c10445589f"},"schema_version":"1.0"},"canonical_sha256":"4ac0ada97322ac477bf9ef6fd27d612268c7dac39e2ee39692b768f17ed901fc","source":{"kind":"arxiv","id":"2502.15815","version":1},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2502.15815","created_at":"2026-07-05T10:18:16Z"},{"alias_kind":"arxiv_version","alias_value":"2502.15815v1","created_at":"2026-07-05T10:18:16Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2502.15815","created_at":"2026-07-05T10:18:16Z"},{"alias_kind":"pith_short_12","alias_value":"JLAK3KLTEKWE","created_at":"2026-07-05T10:18:16Z"},{"alias_kind":"pith_short_16","alias_value":"JLAK3KLTEKWEO67Z","created_at":"2026-07-05T10:18:16Z"},{"alias_kind":"pith_short_8","alias_value":"JLAK3KLT","created_at":"2026-07-05T10:18:16Z"}],"events":[{"event_type":"record_created","subject_pith_number":"pith:2025:JLAK3KLTEKWEO67Z55X5E7LBEJ","target":"record","payload":{"canonical_record":{"source":{"id":"2502.15815","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.LG","submitted_at":"2025-02-19T19:00:00Z","cross_cats_sorted":["astro-ph.CO","cs.AI","hep-ph","hep-th"],"title_canon_sha256":"5b93d434733e2ae2b5b4c9b6be2f928080ed23b1fd5d817fd6b594a8d4ed630c","abstract_canon_sha256":"956a41971638ae3a578f5da035c1ba690d2f85fd9befa7994af120c10445589f"},"schema_version":"1.0"},"canonical_sha256":"4ac0ada97322ac477bf9ef6fd27d612268c7dac39e2ee39692b768f17ed901fc","receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T10:18:16.739639Z","signature_b64":"3Yfjl5hNjar53kYXg8jvXoEp9DJnanPoXVjJ1hg5STT7D+SwROzQ18AhOjcleEhAko8LUKGfD+Tte68gHa8eAA==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"4ac0ada97322ac477bf9ef6fd27d612268c7dac39e2ee39692b768f17ed901fc","last_reissued_at":"2026-07-05T10:18:16.739204Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T10:18:16.739204Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"source_kind":"arxiv","source_id":"2502.15815","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":"2026-07-05T10:18:16Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"rW0XGhEVq+M/qD2IY1lpnt0BL1UKXIDCi217fHo9pxLSkkH8gubtz8CHViXZeSBr8aip7dIjG2hnJxsKDrJfAg==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-08-21T06:14:07.137878Z"},"content_sha256":"a87c251604c62605dccab5d9627eedc5dc784279bb51618c1236a730bc08bf89","schema_version":"1.0","event_id":"sha256:a87c251604c62605dccab5d9627eedc5dc784279bb51618c1236a730bc08bf89"},{"event_type":"graph_snapshot","subject_pith_number":"pith:2025:JLAK3KLTEKWEO67Z55X5E7LBEJ","target":"graph","payload":{"graph_snapshot":{"paper":{"title":"Theoretical Physics Benchmark (TPBench) -- a Dataset and Study of AI Reasoning Capabilities in Theoretical Physics","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["astro-ph.CO","cs.AI","hep-ph","hep-th"],"primary_cat":"cs.LG","authors_text":"Daniel J.H. Chung, Frederic Sala, Maja Rudolph, Moritz M\\\"unchmeyer, Sai Chaitanya Tadepalli, Tianyi Li, Yurii Kvasiuk, Zhiqi Gao","submitted_at":"2025-02-19T19:00:00Z","abstract_excerpt":"We introduce a benchmark to evaluate the capability of AI to solve problems in theoretical physics, focusing on high-energy theory and cosmology. The first iteration of our benchmark consists of 57 problems of varying difficulty, from undergraduate to research level. These problems are novel in the sense that they do not come from public problem collections. We evaluate our data set on various open and closed language models, including o3-mini, o1, DeepSeek-R1, GPT-4o and versions of Llama and Qwen. While we find impressive progress in model performance with the most recent models, our researc"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2502.15815","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/2502.15815/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":"2026-07-05T10:18:16Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"aQ1yzN29HucuXRBQ06JcVLm5Qwu0VLjSbCKBDQ+yJ9ztkAbaOvX/KigLu1QpckGXdJnfMdBX9pw26WDeADTrDA==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-08-21T06:14:07.138412Z"},"content_sha256":"9114916fdcd540fb373eee608135521eb1831d618988241ba762edca9d16ed59","schema_version":"1.0","event_id":"sha256:9114916fdcd540fb373eee608135521eb1831d618988241ba762edca9d16ed59"}],"timestamp_proofs":[],"mirror_hints":[{"mirror_type":"https","name":"Pith Resolver","base_url":"https://pith.science","bundle_url":"https://pith.science/pith/JLAK3KLTEKWEO67Z55X5E7LBEJ/bundle.json","state_url":"https://pith.science/pith/JLAK3KLTEKWEO67Z55X5E7LBEJ/state.json","well_known_bundle_url":"https://pith.science/.well-known/pith/JLAK3KLTEKWEO67Z55X5E7LBEJ/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":"2026-08-21T06:14:07Z","links":{"resolver":"https://pith.science/pith/JLAK3KLTEKWEO67Z55X5E7LBEJ","bundle":"https://pith.science/pith/JLAK3KLTEKWEO67Z55X5E7LBEJ/bundle.json","state":"https://pith.science/pith/JLAK3KLTEKWEO67Z55X5E7LBEJ/state.json","well_known_bundle":"https://pith.science/.well-known/pith/JLAK3KLTEKWEO67Z55X5E7LBEJ/bundle.json"},"state":{"state_type":"pith_open_graph_state","state_version":"1.0","pith_number":"pith:2025:JLAK3KLTEKWEO67Z55X5E7LBEJ","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":"956a41971638ae3a578f5da035c1ba690d2f85fd9befa7994af120c10445589f","cross_cats_sorted":["astro-ph.CO","cs.AI","hep-ph","hep-th"],"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.LG","submitted_at":"2025-02-19T19:00:00Z","title_canon_sha256":"5b93d434733e2ae2b5b4c9b6be2f928080ed23b1fd5d817fd6b594a8d4ed630c"},"schema_version":"1.0","source":{"id":"2502.15815","kind":"arxiv","version":1}},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2502.15815","created_at":"2026-07-05T10:18:16Z"},{"alias_kind":"arxiv_version","alias_value":"2502.15815v1","created_at":"2026-07-05T10:18:16Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2502.15815","created_at":"2026-07-05T10:18:16Z"},{"alias_kind":"pith_short_12","alias_value":"JLAK3KLTEKWE","created_at":"2026-07-05T10:18:16Z"},{"alias_kind":"pith_short_16","alias_value":"JLAK3KLTEKWEO67Z","created_at":"2026-07-05T10:18:16Z"},{"alias_kind":"pith_short_8","alias_value":"JLAK3KLT","created_at":"2026-07-05T10:18:16Z"}],"graph_snapshots":[{"event_id":"sha256:9114916fdcd540fb373eee608135521eb1831d618988241ba762edca9d16ed59","target":"graph","created_at":"2026-07-05T10:18:16Z","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/2502.15815/integrity.json","findings":[],"snapshot_sha256":"c28c3603d3b5d939e8dc4c7e95fa8dfce3d595e45f758748cecf8e644a296938","summary":{"advisory":0,"by_detector":{},"critical":0,"informational":0}},"paper":{"abstract_excerpt":"We introduce a benchmark to evaluate the capability of AI to solve problems in theoretical physics, focusing on high-energy theory and cosmology. The first iteration of our benchmark consists of 57 problems of varying difficulty, from undergraduate to research level. These problems are novel in the sense that they do not come from public problem collections. We evaluate our data set on various open and closed language models, including o3-mini, o1, DeepSeek-R1, GPT-4o and versions of Llama and Qwen. While we find impressive progress in model performance with the most recent models, our researc","authors_text":"Daniel J.H. Chung, Frederic Sala, Maja Rudolph, Moritz M\\\"unchmeyer, Sai Chaitanya Tadepalli, Tianyi Li, Yurii Kvasiuk, Zhiqi Gao","cross_cats":["astro-ph.CO","cs.AI","hep-ph","hep-th"],"headline":"","license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.LG","submitted_at":"2025-02-19T19:00:00Z","title":"Theoretical Physics Benchmark (TPBench) -- a Dataset and Study of AI Reasoning Capabilities in Theoretical Physics"},"references":{"count":0,"internal_anchors":0,"resolved_work":0,"sample":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2502.15815","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:a87c251604c62605dccab5d9627eedc5dc784279bb51618c1236a730bc08bf89","target":"record","created_at":"2026-07-05T10:18:16Z","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":"956a41971638ae3a578f5da035c1ba690d2f85fd9befa7994af120c10445589f","cross_cats_sorted":["astro-ph.CO","cs.AI","hep-ph","hep-th"],"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.LG","submitted_at":"2025-02-19T19:00:00Z","title_canon_sha256":"5b93d434733e2ae2b5b4c9b6be2f928080ed23b1fd5d817fd6b594a8d4ed630c"},"schema_version":"1.0","source":{"id":"2502.15815","kind":"arxiv","version":1}},"canonical_sha256":"4ac0ada97322ac477bf9ef6fd27d612268c7dac39e2ee39692b768f17ed901fc","receipt":{"algorithm":"ed25519","builder_version":"pith-number-builder-2026-05-17-v1","canonical_sha256":"4ac0ada97322ac477bf9ef6fd27d612268c7dac39e2ee39692b768f17ed901fc","first_computed_at":"2026-07-05T10:18:16.739204Z","key_id":"pith-v1-2026-05","kind":"pith_receipt","last_reissued_at":"2026-07-05T10:18:16.739204Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","receipt_version":"0.3","signature_b64":"3Yfjl5hNjar53kYXg8jvXoEp9DJnanPoXVjJ1hg5STT7D+SwROzQ18AhOjcleEhAko8LUKGfD+Tte68gHa8eAA==","signature_status":"signed_v1","signed_at":"2026-07-05T10:18:16.739639Z","signed_message":"canonical_sha256_bytes"},"source_id":"2502.15815","source_kind":"arxiv","source_version":1}}},"equivocations":[],"invalid_events":[],"applied_event_ids":["sha256:a87c251604c62605dccab5d9627eedc5dc784279bb51618c1236a730bc08bf89","sha256:9114916fdcd540fb373eee608135521eb1831d618988241ba762edca9d16ed59"],"state_sha256":"5f7927eb7306338494bf91366105dc6a97d394f99f90e3c2affae34018e01d7e"},"bundle_signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"eKWKMk83NEOHMgufFsLxiLBpMxGj8HDbe97IhkeG7kQOW5uc9LTahvHt3a2TMuQ0JI6kJCZQmGAp13+wKBSUBA==","signed_message":"bundle_sha256_bytes","signed_at":"2026-08-21T06:14:07.142988Z","bundle_sha256":"14f42f913ed8058e3199117728be2a634aa5f2b43ce95869b63c5976131abdcd"}}