{"bundle_type":"pith_open_graph_bundle","bundle_version":"1.0","pith_number":"pith:2025:4I63GVLR4B5YOG3TU36R227RMD","short_pith_number":"pith:4I63GVLR","canonical_record":{"source":{"id":"2502.06738","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.LG","submitted_at":"2025-02-10T18:07:09Z","cross_cats_sorted":[],"title_canon_sha256":"bb6dbb7b2ec89cdf84309dbde9529d8205867068cb9bc1aae9bc88fe04039a29","abstract_canon_sha256":"b6e6342cd2390c91f873a39127458512c562be49154399bead5e53ddd609a465"},"schema_version":"1.0"},"canonical_sha256":"e23db35571e07b871b73a6fd1d6bf160e75076063027862cde6f21eed32c238b","source":{"kind":"arxiv","id":"2502.06738","version":1},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2502.06738","created_at":"2026-07-05T10:12:11Z"},{"alias_kind":"arxiv_version","alias_value":"2502.06738v1","created_at":"2026-07-05T10:12:11Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2502.06738","created_at":"2026-07-05T10:12:11Z"},{"alias_kind":"pith_short_12","alias_value":"4I63GVLR4B5Y","created_at":"2026-07-05T10:12:11Z"},{"alias_kind":"pith_short_16","alias_value":"4I63GVLR4B5YOG3T","created_at":"2026-07-05T10:12:11Z"},{"alias_kind":"pith_short_8","alias_value":"4I63GVLR","created_at":"2026-07-05T10:12:11Z"}],"events":[{"event_type":"record_created","subject_pith_number":"pith:2025:4I63GVLR4B5YOG3TU36R227RMD","target":"record","payload":{"canonical_record":{"source":{"id":"2502.06738","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.LG","submitted_at":"2025-02-10T18:07:09Z","cross_cats_sorted":[],"title_canon_sha256":"bb6dbb7b2ec89cdf84309dbde9529d8205867068cb9bc1aae9bc88fe04039a29","abstract_canon_sha256":"b6e6342cd2390c91f873a39127458512c562be49154399bead5e53ddd609a465"},"schema_version":"1.0"},"canonical_sha256":"e23db35571e07b871b73a6fd1d6bf160e75076063027862cde6f21eed32c238b","receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T10:12:11.523187Z","signature_b64":"r6EsmPZ27xRPfBy9Xhq63IBWRnfrxaFPJFx+qprUP/pyXrmRkQIdAZ5lP8NUPeg3FhJ0ULKjr8GFA0G/u9doAg==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"e23db35571e07b871b73a6fd1d6bf160e75076063027862cde6f21eed32c238b","last_reissued_at":"2026-07-05T10:12:11.522732Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T10:12:11.522732Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"source_kind":"arxiv","source_id":"2502.06738","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":"2026-07-05T10:12:11Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"OElFcNB5FUQi5lo+p9klGWjbtyjvRlZ1u3V+Fuf03k+lLgmaDAxDhHP8ialMs6yc0zqMqHuO7yDxfMOkR233AA==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-08-10T05:54:35.669621Z"},"content_sha256":"c5613c9e5e06917f703316b8ac3f812c18318e3b1a0ffce5e971fbddd813ae84","schema_version":"1.0","event_id":"sha256:c5613c9e5e06917f703316b8ac3f812c18318e3b1a0ffce5e971fbddd813ae84"},{"event_type":"graph_snapshot","subject_pith_number":"pith:2025:4I63GVLR4B5YOG3TU36R227RMD","target":"graph","payload":{"graph_snapshot":{"paper":{"title":"Resurrecting saturated LLM benchmarks with adversarial encoding","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":[],"primary_cat":"cs.LG","authors_text":"Dmitrii Volkov, Igor Ivanov","submitted_at":"2025-02-10T18:07:09Z","abstract_excerpt":"Recent work showed that small changes in benchmark questions can reduce LLMs' reasoning and recall. We explore two such changes: pairing questions and adding more answer options, on three benchmarks: WMDP-bio, GPQA, and MMLU variants. We find that for more capable models, these predictably reduce performance, essentially heightening the performance ceiling of a benchmark and unsaturating it again. We suggest this approach can resurrect old benchmarks."},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2502.06738","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/2502.06738/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":"2026-07-05T10:12:11Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"0Ff46xRWQkQKW2AQJ1LLzZEKwYaAKetuvS+Q/opLaU7MXPrVNQPQCPn187Fu0yLjK4ZvN1VJCptUNewSFWgLCA==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-08-10T05:54:35.670490Z"},"content_sha256":"6ffcb5a32e019838c9c3ae9175449c4ef1e4cfd1eec11ba8d5b9736b6884c866","schema_version":"1.0","event_id":"sha256:6ffcb5a32e019838c9c3ae9175449c4ef1e4cfd1eec11ba8d5b9736b6884c866"}],"timestamp_proofs":[],"mirror_hints":[{"mirror_type":"https","name":"Pith Resolver","base_url":"https://pith.science","bundle_url":"https://pith.science/pith/4I63GVLR4B5YOG3TU36R227RMD/bundle.json","state_url":"https://pith.science/pith/4I63GVLR4B5YOG3TU36R227RMD/state.json","well_known_bundle_url":"https://pith.science/.well-known/pith/4I63GVLR4B5YOG3TU36R227RMD/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":"2026-08-10T05:54:35Z","links":{"resolver":"https://pith.science/pith/4I63GVLR4B5YOG3TU36R227RMD","bundle":"https://pith.science/pith/4I63GVLR4B5YOG3TU36R227RMD/bundle.json","state":"https://pith.science/pith/4I63GVLR4B5YOG3TU36R227RMD/state.json","well_known_bundle":"https://pith.science/.well-known/pith/4I63GVLR4B5YOG3TU36R227RMD/bundle.json"},"state":{"state_type":"pith_open_graph_state","state_version":"1.0","pith_number":"pith:2025:4I63GVLR4B5YOG3TU36R227RMD","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":"b6e6342cd2390c91f873a39127458512c562be49154399bead5e53ddd609a465","cross_cats_sorted":[],"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.LG","submitted_at":"2025-02-10T18:07:09Z","title_canon_sha256":"bb6dbb7b2ec89cdf84309dbde9529d8205867068cb9bc1aae9bc88fe04039a29"},"schema_version":"1.0","source":{"id":"2502.06738","kind":"arxiv","version":1}},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2502.06738","created_at":"2026-07-05T10:12:11Z"},{"alias_kind":"arxiv_version","alias_value":"2502.06738v1","created_at":"2026-07-05T10:12:11Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2502.06738","created_at":"2026-07-05T10:12:11Z"},{"alias_kind":"pith_short_12","alias_value":"4I63GVLR4B5Y","created_at":"2026-07-05T10:12:11Z"},{"alias_kind":"pith_short_16","alias_value":"4I63GVLR4B5YOG3T","created_at":"2026-07-05T10:12:11Z"},{"alias_kind":"pith_short_8","alias_value":"4I63GVLR","created_at":"2026-07-05T10:12:11Z"}],"graph_snapshots":[{"event_id":"sha256:6ffcb5a32e019838c9c3ae9175449c4ef1e4cfd1eec11ba8d5b9736b6884c866","target":"graph","created_at":"2026-07-05T10:12:11Z","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/2502.06738/integrity.json","findings":[],"snapshot_sha256":"c28c3603d3b5d939e8dc4c7e95fa8dfce3d595e45f758748cecf8e644a296938","summary":{"advisory":0,"by_detector":{},"critical":0,"informational":0}},"paper":{"abstract_excerpt":"Recent work showed that small changes in benchmark questions can reduce LLMs' reasoning and recall. We explore two such changes: pairing questions and adding more answer options, on three benchmarks: WMDP-bio, GPQA, and MMLU variants. We find that for more capable models, these predictably reduce performance, essentially heightening the performance ceiling of a benchmark and unsaturating it again. We suggest this approach can resurrect old benchmarks.","authors_text":"Dmitrii Volkov, Igor Ivanov","cross_cats":[],"headline":"","license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.LG","submitted_at":"2025-02-10T18:07:09Z","title":"Resurrecting saturated LLM benchmarks with adversarial encoding"},"references":{"count":0,"internal_anchors":0,"resolved_work":0,"sample":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2502.06738","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:c5613c9e5e06917f703316b8ac3f812c18318e3b1a0ffce5e971fbddd813ae84","target":"record","created_at":"2026-07-05T10:12:11Z","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":"b6e6342cd2390c91f873a39127458512c562be49154399bead5e53ddd609a465","cross_cats_sorted":[],"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.LG","submitted_at":"2025-02-10T18:07:09Z","title_canon_sha256":"bb6dbb7b2ec89cdf84309dbde9529d8205867068cb9bc1aae9bc88fe04039a29"},"schema_version":"1.0","source":{"id":"2502.06738","kind":"arxiv","version":1}},"canonical_sha256":"e23db35571e07b871b73a6fd1d6bf160e75076063027862cde6f21eed32c238b","receipt":{"algorithm":"ed25519","builder_version":"pith-number-builder-2026-05-17-v1","canonical_sha256":"e23db35571e07b871b73a6fd1d6bf160e75076063027862cde6f21eed32c238b","first_computed_at":"2026-07-05T10:12:11.522732Z","key_id":"pith-v1-2026-05","kind":"pith_receipt","last_reissued_at":"2026-07-05T10:12:11.522732Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","receipt_version":"0.3","signature_b64":"r6EsmPZ27xRPfBy9Xhq63IBWRnfrxaFPJFx+qprUP/pyXrmRkQIdAZ5lP8NUPeg3FhJ0ULKjr8GFA0G/u9doAg==","signature_status":"signed_v1","signed_at":"2026-07-05T10:12:11.523187Z","signed_message":"canonical_sha256_bytes"},"source_id":"2502.06738","source_kind":"arxiv","source_version":1}}},"equivocations":[],"invalid_events":[],"applied_event_ids":["sha256:c5613c9e5e06917f703316b8ac3f812c18318e3b1a0ffce5e971fbddd813ae84","sha256:6ffcb5a32e019838c9c3ae9175449c4ef1e4cfd1eec11ba8d5b9736b6884c866"],"state_sha256":"c24fdfa679b9f6766d9031d224e88a72596072ffbe60c0a0169b4be4550c83ac"},"bundle_signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"du3MLPYn3zDSmjq9JG3y6NJ1iG4CtnjeEEpmDZZmH+O4f7VaBMTTs84iTD6/oKZ/Lm/zMw0uQYxVO4S3pf0fDQ==","signed_message":"bundle_sha256_bytes","signed_at":"2026-08-10T05:54:35.676352Z","bundle_sha256":"e412fa3c59a2102bb3cb30a27241f328d3c3c7e81a6df80ff5ac85583df5cde5"}}