{"bundle_type":"pith_open_graph_bundle","bundle_version":"1.0","pith_number":"pith:2023:VLEYFIMZ2R36XZMLXZQCMN2Q4G","short_pith_number":"pith:VLEYFIMZ","canonical_record":{"source":{"id":"2312.13547","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CL","submitted_at":"2023-12-21T03:11:30Z","cross_cats_sorted":[],"title_canon_sha256":"96ef400058108177db93f2d71cbd7740e3f02aecab5f813d238286f09bf3aa20","abstract_canon_sha256":"5a0cd849383a7a0900b767db94273530f22e4a22aed8e76df47399b3f3a51033"},"schema_version":"1.0"},"canonical_sha256":"aac982a199d477ebe58bbe60263750e1b81d05bdadaa0183be186545b8a2abc8","source":{"kind":"arxiv","id":"2312.13547","version":1},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2312.13547","created_at":"2026-07-05T07:26:46Z"},{"alias_kind":"arxiv_version","alias_value":"2312.13547v1","created_at":"2026-07-05T07:26:46Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2312.13547","created_at":"2026-07-05T07:26:46Z"},{"alias_kind":"pith_short_12","alias_value":"VLEYFIMZ2R36","created_at":"2026-07-05T07:26:46Z"},{"alias_kind":"pith_short_16","alias_value":"VLEYFIMZ2R36XZML","created_at":"2026-07-05T07:26:46Z"},{"alias_kind":"pith_short_8","alias_value":"VLEYFIMZ","created_at":"2026-07-05T07:26:46Z"}],"events":[{"event_type":"record_created","subject_pith_number":"pith:2023:VLEYFIMZ2R36XZMLXZQCMN2Q4G","target":"record","payload":{"canonical_record":{"source":{"id":"2312.13547","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CL","submitted_at":"2023-12-21T03:11:30Z","cross_cats_sorted":[],"title_canon_sha256":"96ef400058108177db93f2d71cbd7740e3f02aecab5f813d238286f09bf3aa20","abstract_canon_sha256":"5a0cd849383a7a0900b767db94273530f22e4a22aed8e76df47399b3f3a51033"},"schema_version":"1.0"},"canonical_sha256":"aac982a199d477ebe58bbe60263750e1b81d05bdadaa0183be186545b8a2abc8","receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T07:26:46.308402Z","signature_b64":"dGNKDzwQt7bD3zQbtBGltUdllikfho8pCHKuGzS2tJxHBcyoCYHWNllJuJ5pKOJT5ZdGRqwJtl5x8BF1KBE9Ag==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"aac982a199d477ebe58bbe60263750e1b81d05bdadaa0183be186545b8a2abc8","last_reissued_at":"2026-07-05T07:26:46.308065Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T07:26:46.308065Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"source_kind":"arxiv","source_id":"2312.13547","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":"2026-07-05T07:26:46Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"zFEmA1uUiuFC4bwEhzc5zdS9Kd807X5VNw8PbAB6oLWn7eshRgPbg+Std8Bnou+ejguQ5+M8UaW+m/msw3p/Bw==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-08-15T11:27:11.518809Z"},"content_sha256":"f54b1b4e6338cfce1eef3b5076c3d8b49ffa3eb0f4fbc9916772c4ae7cbcc7cc","schema_version":"1.0","event_id":"sha256:f54b1b4e6338cfce1eef3b5076c3d8b49ffa3eb0f4fbc9916772c4ae7cbcc7cc"},{"event_type":"graph_snapshot","subject_pith_number":"pith:2023:VLEYFIMZ2R36XZMLXZQCMN2Q4G","target":"graph","payload":{"graph_snapshot":{"paper":{"title":"How to Prune Your Language Model: Recovering Accuracy on the \"Sparsity May Cry'' Benchmark","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":[],"primary_cat":"cs.CL","authors_text":"Dan Alistarh, Eldar Kurtic, Torsten Hoefler","submitted_at":"2023-12-21T03:11:30Z","abstract_excerpt":"Pruning large language models (LLMs) from the BERT family has emerged as a standard compression benchmark, and several pruning methods have been proposed for this task. The recent ``Sparsity May Cry'' (SMC) benchmark put into question the validity of all existing methods, exhibiting a more complex setup where many known pruning methods appear to fail. We revisit the question of accurate BERT-pruning during fine-tuning on downstream datasets, and propose a set of general guidelines for successful pruning, even on the challenging SMC benchmark. First, we perform a cost-vs-benefits analysis of pr"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2312.13547","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/2312.13547/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":"2026-07-05T07:26:46Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"2rZ4rVSqht3ySHHhm1+pVTth1jMwWm3clKTzvrEi/yguPMOELONeext9KD1LVhDJjy1lwDzP6ztK0y951R09CA==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-08-15T11:27:11.519397Z"},"content_sha256":"7952cf67f3781d50569d5bfd498023a2bc2ef030fe8713d0683aa70f0bfe71ba","schema_version":"1.0","event_id":"sha256:7952cf67f3781d50569d5bfd498023a2bc2ef030fe8713d0683aa70f0bfe71ba"}],"timestamp_proofs":[],"mirror_hints":[{"mirror_type":"https","name":"Pith Resolver","base_url":"https://pith.science","bundle_url":"https://pith.science/pith/VLEYFIMZ2R36XZMLXZQCMN2Q4G/bundle.json","state_url":"https://pith.science/pith/VLEYFIMZ2R36XZMLXZQCMN2Q4G/state.json","well_known_bundle_url":"https://pith.science/.well-known/pith/VLEYFIMZ2R36XZMLXZQCMN2Q4G/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":"2026-08-15T11:27:11Z","links":{"resolver":"https://pith.science/pith/VLEYFIMZ2R36XZMLXZQCMN2Q4G","bundle":"https://pith.science/pith/VLEYFIMZ2R36XZMLXZQCMN2Q4G/bundle.json","state":"https://pith.science/pith/VLEYFIMZ2R36XZMLXZQCMN2Q4G/state.json","well_known_bundle":"https://pith.science/.well-known/pith/VLEYFIMZ2R36XZMLXZQCMN2Q4G/bundle.json"},"state":{"state_type":"pith_open_graph_state","state_version":"1.0","pith_number":"pith:2023:VLEYFIMZ2R36XZMLXZQCMN2Q4G","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":"5a0cd849383a7a0900b767db94273530f22e4a22aed8e76df47399b3f3a51033","cross_cats_sorted":[],"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CL","submitted_at":"2023-12-21T03:11:30Z","title_canon_sha256":"96ef400058108177db93f2d71cbd7740e3f02aecab5f813d238286f09bf3aa20"},"schema_version":"1.0","source":{"id":"2312.13547","kind":"arxiv","version":1}},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2312.13547","created_at":"2026-07-05T07:26:46Z"},{"alias_kind":"arxiv_version","alias_value":"2312.13547v1","created_at":"2026-07-05T07:26:46Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2312.13547","created_at":"2026-07-05T07:26:46Z"},{"alias_kind":"pith_short_12","alias_value":"VLEYFIMZ2R36","created_at":"2026-07-05T07:26:46Z"},{"alias_kind":"pith_short_16","alias_value":"VLEYFIMZ2R36XZML","created_at":"2026-07-05T07:26:46Z"},{"alias_kind":"pith_short_8","alias_value":"VLEYFIMZ","created_at":"2026-07-05T07:26:46Z"}],"graph_snapshots":[{"event_id":"sha256:7952cf67f3781d50569d5bfd498023a2bc2ef030fe8713d0683aa70f0bfe71ba","target":"graph","created_at":"2026-07-05T07:26:46Z","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/2312.13547/integrity.json","findings":[],"snapshot_sha256":"c28c3603d3b5d939e8dc4c7e95fa8dfce3d595e45f758748cecf8e644a296938","summary":{"advisory":0,"by_detector":{},"critical":0,"informational":0}},"paper":{"abstract_excerpt":"Pruning large language models (LLMs) from the BERT family has emerged as a standard compression benchmark, and several pruning methods have been proposed for this task. The recent ``Sparsity May Cry'' (SMC) benchmark put into question the validity of all existing methods, exhibiting a more complex setup where many known pruning methods appear to fail. We revisit the question of accurate BERT-pruning during fine-tuning on downstream datasets, and propose a set of general guidelines for successful pruning, even on the challenging SMC benchmark. First, we perform a cost-vs-benefits analysis of pr","authors_text":"Dan Alistarh, Eldar Kurtic, Torsten Hoefler","cross_cats":[],"headline":"","license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CL","submitted_at":"2023-12-21T03:11:30Z","title":"How to Prune Your Language Model: Recovering Accuracy on the \"Sparsity May Cry'' Benchmark"},"references":{"count":0,"internal_anchors":0,"resolved_work":0,"sample":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2312.13547","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:f54b1b4e6338cfce1eef3b5076c3d8b49ffa3eb0f4fbc9916772c4ae7cbcc7cc","target":"record","created_at":"2026-07-05T07:26:46Z","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":"5a0cd849383a7a0900b767db94273530f22e4a22aed8e76df47399b3f3a51033","cross_cats_sorted":[],"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CL","submitted_at":"2023-12-21T03:11:30Z","title_canon_sha256":"96ef400058108177db93f2d71cbd7740e3f02aecab5f813d238286f09bf3aa20"},"schema_version":"1.0","source":{"id":"2312.13547","kind":"arxiv","version":1}},"canonical_sha256":"aac982a199d477ebe58bbe60263750e1b81d05bdadaa0183be186545b8a2abc8","receipt":{"algorithm":"ed25519","builder_version":"pith-number-builder-2026-05-17-v1","canonical_sha256":"aac982a199d477ebe58bbe60263750e1b81d05bdadaa0183be186545b8a2abc8","first_computed_at":"2026-07-05T07:26:46.308065Z","key_id":"pith-v1-2026-05","kind":"pith_receipt","last_reissued_at":"2026-07-05T07:26:46.308065Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","receipt_version":"0.3","signature_b64":"dGNKDzwQt7bD3zQbtBGltUdllikfho8pCHKuGzS2tJxHBcyoCYHWNllJuJ5pKOJT5ZdGRqwJtl5x8BF1KBE9Ag==","signature_status":"signed_v1","signed_at":"2026-07-05T07:26:46.308402Z","signed_message":"canonical_sha256_bytes"},"source_id":"2312.13547","source_kind":"arxiv","source_version":1}}},"equivocations":[],"invalid_events":[],"applied_event_ids":["sha256:f54b1b4e6338cfce1eef3b5076c3d8b49ffa3eb0f4fbc9916772c4ae7cbcc7cc","sha256:7952cf67f3781d50569d5bfd498023a2bc2ef030fe8713d0683aa70f0bfe71ba"],"state_sha256":"9dddb044af26297ad4e4ad43e2f8e9dbd790f11aba10b2619679ca158928ec43"},"bundle_signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"M8CLxt4Fm7ovXuyYzalNDuExQa+Wsf5zaVQD7Xf+4Cikc23ZzJJgGzN4Zlcg4MtV7kb0R3l7quedKLYYgtyvBQ==","signed_message":"bundle_sha256_bytes","signed_at":"2026-08-15T11:27:11.525573Z","bundle_sha256":"6c6e91d4ee8f043770ad6954da2c864daf8f71ea1264c42be41d33ce2951c557"}}