{"bundle_type":"pith_open_graph_bundle","bundle_version":"1.0","pith_number":"pith:2024:TZBCUBE5MI2CRNOGYZ5LHIPDFZ","short_pith_number":"pith:TZBCUBE5","canonical_record":{"source":{"id":"2403.00393","kind":"arxiv","version":2},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CR","submitted_at":"2024-03-01T09:28:38Z","cross_cats_sorted":["cs.CL"],"title_canon_sha256":"c8191ba05a52bdb2f98eafd36e6f1b3d460a2910ccd3855938ddd9d34266e522","abstract_canon_sha256":"a1b456d1463e22b94b3787dfcfaff1431b75758775d623a51cb684796883299f"},"schema_version":"1.0"},"canonical_sha256":"9e422a049d623428b5c6c67ab3a1e32e671647056760d7f423d728162d7de65f","source":{"kind":"arxiv","id":"2403.00393","version":2},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2403.00393","created_at":"2026-07-05T08:35:43Z"},{"alias_kind":"arxiv_version","alias_value":"2403.00393v2","created_at":"2026-07-05T08:35:43Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2403.00393","created_at":"2026-07-05T08:35:43Z"},{"alias_kind":"pith_short_12","alias_value":"TZBCUBE5MI2C","created_at":"2026-07-05T08:35:43Z"},{"alias_kind":"pith_short_16","alias_value":"TZBCUBE5MI2CRNOG","created_at":"2026-07-05T08:35:43Z"},{"alias_kind":"pith_short_8","alias_value":"TZBCUBE5","created_at":"2026-07-05T08:35:43Z"}],"events":[{"event_type":"record_created","subject_pith_number":"pith:2024:TZBCUBE5MI2CRNOGYZ5LHIPDFZ","target":"record","payload":{"canonical_record":{"source":{"id":"2403.00393","kind":"arxiv","version":2},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CR","submitted_at":"2024-03-01T09:28:38Z","cross_cats_sorted":["cs.CL"],"title_canon_sha256":"c8191ba05a52bdb2f98eafd36e6f1b3d460a2910ccd3855938ddd9d34266e522","abstract_canon_sha256":"a1b456d1463e22b94b3787dfcfaff1431b75758775d623a51cb684796883299f"},"schema_version":"1.0"},"canonical_sha256":"9e422a049d623428b5c6c67ab3a1e32e671647056760d7f423d728162d7de65f","receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T08:35:43.015344Z","signature_b64":"yw5cce0RXBOfKO864/nXyLA1zNLIhznXmjxkPwni7onaw1UDlbqsuhFQasE+A3By5+da3JxPC1NsR+2dxaO9Ag==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"9e422a049d623428b5c6c67ab3a1e32e671647056760d7f423d728162d7de65f","last_reissued_at":"2026-07-05T08:35:43.014927Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T08:35:43.014927Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"source_kind":"arxiv","source_id":"2403.00393","source_version":2,"attestation_state":"computed"},"signer":{"signer_id":"pith.science","signer_type":"pith_registry","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"created_at":"2026-07-05T08:35:43Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"csHBn9JWyD6yyZJ22lHjsgKp7xt75vngXed1jaqA5r6AZkFIOMjvGCm4q7Raku1nSp2go7sdWq5Vgtlz7B73Ag==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-07-07T04:16:55.053100Z"},"content_sha256":"123583e1067b7c5c19cf6146220902799ebb30707745c00713120aeb795a5dc3","schema_version":"1.0","event_id":"sha256:123583e1067b7c5c19cf6146220902799ebb30707745c00713120aeb795a5dc3"},{"event_type":"graph_snapshot","subject_pith_number":"pith:2024:TZBCUBE5MI2CRNOGYZ5LHIPDFZ","target":"graph","payload":{"graph_snapshot":{"paper":{"title":"TRUCE: Private Benchmarking to Prevent Contamination and Improve Comparative Evaluation of LLMs","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.CL"],"primary_cat":"cs.CR","authors_text":"Divya Gupta, Kashish Mittal, Manohar Swaminathan, Nishanth Chandran, Rahul Sharma, Sunayana Sitaram, Tanmay Rajore","submitted_at":"2024-03-01T09:28:38Z","abstract_excerpt":"Benchmarking is the de-facto standard for evaluating LLMs, due to its speed, replicability and low cost. However, recent work has pointed out that the majority of the open source benchmarks available today have been contaminated or leaked into LLMs, meaning that LLMs have access to test data during pretraining and/or fine-tuning. This raises serious concerns about the validity of benchmarking studies conducted so far and the future of evaluation using benchmarks. To solve this problem, we propose Private Benchmarking, a solution where test datasets are kept private and models are evaluated wit"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2403.00393","kind":"arxiv","version":2},"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/2403.00393/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":"2026-07-05T08:35:43Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"7gwxs0r9ZCkCVA1rMQFLW1Y+H1Z2PfWiux1Y3fiAH1MQYaDcntlfaZkfs9EtZf+lhus2HU/nLAd4ibP0f3NaDg==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-07-07T04:16:55.053506Z"},"content_sha256":"96625cc451175e475d19473568f8099b1bb649623aa088009c35f8281785b883","schema_version":"1.0","event_id":"sha256:96625cc451175e475d19473568f8099b1bb649623aa088009c35f8281785b883"}],"timestamp_proofs":[],"mirror_hints":[{"mirror_type":"https","name":"Pith Resolver","base_url":"https://pith.science","bundle_url":"https://pith.science/pith/TZBCUBE5MI2CRNOGYZ5LHIPDFZ/bundle.json","state_url":"https://pith.science/pith/TZBCUBE5MI2CRNOGYZ5LHIPDFZ/state.json","well_known_bundle_url":"https://pith.science/.well-known/pith/TZBCUBE5MI2CRNOGYZ5LHIPDFZ/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":"2026-07-07T04:16:55Z","links":{"resolver":"https://pith.science/pith/TZBCUBE5MI2CRNOGYZ5LHIPDFZ","bundle":"https://pith.science/pith/TZBCUBE5MI2CRNOGYZ5LHIPDFZ/bundle.json","state":"https://pith.science/pith/TZBCUBE5MI2CRNOGYZ5LHIPDFZ/state.json","well_known_bundle":"https://pith.science/.well-known/pith/TZBCUBE5MI2CRNOGYZ5LHIPDFZ/bundle.json"},"state":{"state_type":"pith_open_graph_state","state_version":"1.0","pith_number":"pith:2024:TZBCUBE5MI2CRNOGYZ5LHIPDFZ","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":"a1b456d1463e22b94b3787dfcfaff1431b75758775d623a51cb684796883299f","cross_cats_sorted":["cs.CL"],"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CR","submitted_at":"2024-03-01T09:28:38Z","title_canon_sha256":"c8191ba05a52bdb2f98eafd36e6f1b3d460a2910ccd3855938ddd9d34266e522"},"schema_version":"1.0","source":{"id":"2403.00393","kind":"arxiv","version":2}},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2403.00393","created_at":"2026-07-05T08:35:43Z"},{"alias_kind":"arxiv_version","alias_value":"2403.00393v2","created_at":"2026-07-05T08:35:43Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2403.00393","created_at":"2026-07-05T08:35:43Z"},{"alias_kind":"pith_short_12","alias_value":"TZBCUBE5MI2C","created_at":"2026-07-05T08:35:43Z"},{"alias_kind":"pith_short_16","alias_value":"TZBCUBE5MI2CRNOG","created_at":"2026-07-05T08:35:43Z"},{"alias_kind":"pith_short_8","alias_value":"TZBCUBE5","created_at":"2026-07-05T08:35:43Z"}],"graph_snapshots":[{"event_id":"sha256:96625cc451175e475d19473568f8099b1bb649623aa088009c35f8281785b883","target":"graph","created_at":"2026-07-05T08:35:43Z","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/2403.00393/integrity.json","findings":[],"snapshot_sha256":"c28c3603d3b5d939e8dc4c7e95fa8dfce3d595e45f758748cecf8e644a296938","summary":{"advisory":0,"by_detector":{},"critical":0,"informational":0}},"paper":{"abstract_excerpt":"Benchmarking is the de-facto standard for evaluating LLMs, due to its speed, replicability and low cost. However, recent work has pointed out that the majority of the open source benchmarks available today have been contaminated or leaked into LLMs, meaning that LLMs have access to test data during pretraining and/or fine-tuning. This raises serious concerns about the validity of benchmarking studies conducted so far and the future of evaluation using benchmarks. To solve this problem, we propose Private Benchmarking, a solution where test datasets are kept private and models are evaluated wit","authors_text":"Divya Gupta, Kashish Mittal, Manohar Swaminathan, Nishanth Chandran, Rahul Sharma, Sunayana Sitaram, Tanmay Rajore","cross_cats":["cs.CL"],"headline":"","license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CR","submitted_at":"2024-03-01T09:28:38Z","title":"TRUCE: Private Benchmarking to Prevent Contamination and Improve Comparative Evaluation of LLMs"},"references":{"count":0,"internal_anchors":0,"resolved_work":0,"sample":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2403.00393","kind":"arxiv","version":2},"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":"sha256:123583e1067b7c5c19cf6146220902799ebb30707745c00713120aeb795a5dc3","target":"record","created_at":"2026-07-05T08:35:43Z","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":"a1b456d1463e22b94b3787dfcfaff1431b75758775d623a51cb684796883299f","cross_cats_sorted":["cs.CL"],"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CR","submitted_at":"2024-03-01T09:28:38Z","title_canon_sha256":"c8191ba05a52bdb2f98eafd36e6f1b3d460a2910ccd3855938ddd9d34266e522"},"schema_version":"1.0","source":{"id":"2403.00393","kind":"arxiv","version":2}},"canonical_sha256":"9e422a049d623428b5c6c67ab3a1e32e671647056760d7f423d728162d7de65f","receipt":{"algorithm":"ed25519","builder_version":"pith-number-builder-2026-05-17-v1","canonical_sha256":"9e422a049d623428b5c6c67ab3a1e32e671647056760d7f423d728162d7de65f","first_computed_at":"2026-07-05T08:35:43.014927Z","key_id":"pith-v1-2026-05","kind":"pith_receipt","last_reissued_at":"2026-07-05T08:35:43.014927Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","receipt_version":"0.3","signature_b64":"yw5cce0RXBOfKO864/nXyLA1zNLIhznXmjxkPwni7onaw1UDlbqsuhFQasE+A3By5+da3JxPC1NsR+2dxaO9Ag==","signature_status":"signed_v1","signed_at":"2026-07-05T08:35:43.015344Z","signed_message":"canonical_sha256_bytes"},"source_id":"2403.00393","source_kind":"arxiv","source_version":2}}},"equivocations":[],"invalid_events":[],"applied_event_ids":["sha256:123583e1067b7c5c19cf6146220902799ebb30707745c00713120aeb795a5dc3","sha256:96625cc451175e475d19473568f8099b1bb649623aa088009c35f8281785b883"],"state_sha256":"94b002da0e8d692076e3bb1f12bc3c790eec8a975b5093d338045cf0b61db38f"},"bundle_signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"N28Cp1mMFuEZAg97autwSuZjBtPxv8vX6zJ9nevLxi+1VXrN9a4XXuNPms65O6mWlFD8UZj92xuF19nJ6CN9BA==","signed_message":"bundle_sha256_bytes","signed_at":"2026-07-07T04:16:55.055854Z","bundle_sha256":"3e1aaa459dd7b9b616432486050db8c84404f184ae309a1f573cf76345c336a7"}}