{"bundle_type":"pith_open_graph_bundle","bundle_version":"1.0","pith_number":"pith:2024:EC3EUFUAFZB6GDUSQL6HR4UQ3V","short_pith_number":"pith:EC3EUFUA","canonical_record":{"source":{"id":"2410.07383","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/licenses/by-nc-nd/4.0/","primary_cat":"cs.CL","submitted_at":"2024-10-09T19:03:52Z","cross_cats_sorted":["cs.AI"],"title_canon_sha256":"518ca6e3587975b4a8e3d374eb5960e927901f852cf2ac4be736cd145b2b1c6a","abstract_canon_sha256":"bc75e3a909c3bf4be6d30e53a934cf0ab6e22feb9c72d4671af052a11d0badc4"},"schema_version":"1.0"},"canonical_sha256":"20b64a16802e43e30e9282fc78f290dd65678a4808e84f6ddc2ed0b256422aa9","source":{"kind":"arxiv","id":"2410.07383","version":1},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2410.07383","created_at":"2026-07-05T09:18:35Z"},{"alias_kind":"arxiv_version","alias_value":"2410.07383v1","created_at":"2026-07-05T09:18:35Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2410.07383","created_at":"2026-07-05T09:18:35Z"},{"alias_kind":"pith_short_12","alias_value":"EC3EUFUAFZB6","created_at":"2026-07-05T09:18:35Z"},{"alias_kind":"pith_short_16","alias_value":"EC3EUFUAFZB6GDUS","created_at":"2026-07-05T09:18:35Z"},{"alias_kind":"pith_short_8","alias_value":"EC3EUFUA","created_at":"2026-07-05T09:18:35Z"}],"events":[{"event_type":"record_created","subject_pith_number":"pith:2024:EC3EUFUAFZB6GDUSQL6HR4UQ3V","target":"record","payload":{"canonical_record":{"source":{"id":"2410.07383","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/licenses/by-nc-nd/4.0/","primary_cat":"cs.CL","submitted_at":"2024-10-09T19:03:52Z","cross_cats_sorted":["cs.AI"],"title_canon_sha256":"518ca6e3587975b4a8e3d374eb5960e927901f852cf2ac4be736cd145b2b1c6a","abstract_canon_sha256":"bc75e3a909c3bf4be6d30e53a934cf0ab6e22feb9c72d4671af052a11d0badc4"},"schema_version":"1.0"},"canonical_sha256":"20b64a16802e43e30e9282fc78f290dd65678a4808e84f6ddc2ed0b256422aa9","receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T09:18:35.444638Z","signature_b64":"h+YGqRzIe6q+GVuLyftYAQ0lKbzvkG4wFG74z7koBXIeBL0WgQUbulwf1nw0DpoeeGPT1slShbjekSc5RZZqAA==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"20b64a16802e43e30e9282fc78f290dd65678a4808e84f6ddc2ed0b256422aa9","last_reissued_at":"2026-07-05T09:18:35.444219Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T09:18:35.444219Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"source_kind":"arxiv","source_id":"2410.07383","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":"2026-07-05T09:18:35Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"0t6UXrCuFDnsvezbHIioCtC/r40SbEM048himUcTVo1eQ4ovM9GFF7kSQfalTvz4I8UCTwue92Ezz/cA/TX2Dw==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-07-10T03:32:05.090873Z"},"content_sha256":"fae5c54afc4f9e6b0f7689519f82c1d157447d789d4edbd366be31f1d9088adb","schema_version":"1.0","event_id":"sha256:fae5c54afc4f9e6b0f7689519f82c1d157447d789d4edbd366be31f1d9088adb"},{"event_type":"graph_snapshot","subject_pith_number":"pith:2024:EC3EUFUAFZB6GDUSQL6HR4UQ3V","target":"graph","payload":{"graph_snapshot":{"paper":{"title":"SparseGrad: A Selective Method for Efficient Fine-tuning of MLP Layers","license":"http://creativecommons.org/licenses/by-nc-nd/4.0/","headline":"","cross_cats":["cs.AI"],"primary_cat":"cs.CL","authors_text":"Alexander Mikhalev, Alexander Panchenko, Anna Rudenko, Gleb Mezentsev, Ivan Oseledets, Viktoriia Chekalina","submitted_at":"2024-10-09T19:03:52Z","abstract_excerpt":"The performance of Transformer models has been enhanced by increasing the number of parameters and the length of the processed text. Consequently, fine-tuning the entire model becomes a memory-intensive process. High-performance methods for parameter-efficient fine-tuning (PEFT) typically work with Attention blocks and often overlook MLP blocks, which contain about half of the model parameters. We propose a new selective PEFT method, namely SparseGrad, that performs well on MLP blocks. We transfer layer gradients to a space where only about 1\\% of the layer's elements remain significant. By co"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2410.07383","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/2410.07383/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":"2026-07-05T09:18:35Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"52HsjiHRdSkpuNAdcFQUsJtTlEpVNMOLfBMgBaSuqhhTa+vNTcu425K4ezNt1F+LpUOX3xQu54LVpDkSdLeqAA==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-07-10T03:32:05.091236Z"},"content_sha256":"ac653d1bef5622c7eb9573dfc7a54d27a3910b2a0ca7447fdede1cf2eed52515","schema_version":"1.0","event_id":"sha256:ac653d1bef5622c7eb9573dfc7a54d27a3910b2a0ca7447fdede1cf2eed52515"}],"timestamp_proofs":[],"mirror_hints":[{"mirror_type":"https","name":"Pith Resolver","base_url":"https://pith.science","bundle_url":"https://pith.science/pith/EC3EUFUAFZB6GDUSQL6HR4UQ3V/bundle.json","state_url":"https://pith.science/pith/EC3EUFUAFZB6GDUSQL6HR4UQ3V/state.json","well_known_bundle_url":"https://pith.science/.well-known/pith/EC3EUFUAFZB6GDUSQL6HR4UQ3V/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":"2026-07-10T03:32:05Z","links":{"resolver":"https://pith.science/pith/EC3EUFUAFZB6GDUSQL6HR4UQ3V","bundle":"https://pith.science/pith/EC3EUFUAFZB6GDUSQL6HR4UQ3V/bundle.json","state":"https://pith.science/pith/EC3EUFUAFZB6GDUSQL6HR4UQ3V/state.json","well_known_bundle":"https://pith.science/.well-known/pith/EC3EUFUAFZB6GDUSQL6HR4UQ3V/bundle.json"},"state":{"state_type":"pith_open_graph_state","state_version":"1.0","pith_number":"pith:2024:EC3EUFUAFZB6GDUSQL6HR4UQ3V","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":"bc75e3a909c3bf4be6d30e53a934cf0ab6e22feb9c72d4671af052a11d0badc4","cross_cats_sorted":["cs.AI"],"license":"http://creativecommons.org/licenses/by-nc-nd/4.0/","primary_cat":"cs.CL","submitted_at":"2024-10-09T19:03:52Z","title_canon_sha256":"518ca6e3587975b4a8e3d374eb5960e927901f852cf2ac4be736cd145b2b1c6a"},"schema_version":"1.0","source":{"id":"2410.07383","kind":"arxiv","version":1}},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2410.07383","created_at":"2026-07-05T09:18:35Z"},{"alias_kind":"arxiv_version","alias_value":"2410.07383v1","created_at":"2026-07-05T09:18:35Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2410.07383","created_at":"2026-07-05T09:18:35Z"},{"alias_kind":"pith_short_12","alias_value":"EC3EUFUAFZB6","created_at":"2026-07-05T09:18:35Z"},{"alias_kind":"pith_short_16","alias_value":"EC3EUFUAFZB6GDUS","created_at":"2026-07-05T09:18:35Z"},{"alias_kind":"pith_short_8","alias_value":"EC3EUFUA","created_at":"2026-07-05T09:18:35Z"}],"graph_snapshots":[{"event_id":"sha256:ac653d1bef5622c7eb9573dfc7a54d27a3910b2a0ca7447fdede1cf2eed52515","target":"graph","created_at":"2026-07-05T09:18:35Z","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/2410.07383/integrity.json","findings":[],"snapshot_sha256":"c28c3603d3b5d939e8dc4c7e95fa8dfce3d595e45f758748cecf8e644a296938","summary":{"advisory":0,"by_detector":{},"critical":0,"informational":0}},"paper":{"abstract_excerpt":"The performance of Transformer models has been enhanced by increasing the number of parameters and the length of the processed text. Consequently, fine-tuning the entire model becomes a memory-intensive process. High-performance methods for parameter-efficient fine-tuning (PEFT) typically work with Attention blocks and often overlook MLP blocks, which contain about half of the model parameters. We propose a new selective PEFT method, namely SparseGrad, that performs well on MLP blocks. We transfer layer gradients to a space where only about 1\\% of the layer's elements remain significant. By co","authors_text":"Alexander Mikhalev, Alexander Panchenko, Anna Rudenko, Gleb Mezentsev, Ivan Oseledets, Viktoriia Chekalina","cross_cats":["cs.AI"],"headline":"","license":"http://creativecommons.org/licenses/by-nc-nd/4.0/","primary_cat":"cs.CL","submitted_at":"2024-10-09T19:03:52Z","title":"SparseGrad: A Selective Method for Efficient Fine-tuning of MLP Layers"},"references":{"count":0,"internal_anchors":0,"resolved_work":0,"sample":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2410.07383","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:fae5c54afc4f9e6b0f7689519f82c1d157447d789d4edbd366be31f1d9088adb","target":"record","created_at":"2026-07-05T09:18:35Z","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":"bc75e3a909c3bf4be6d30e53a934cf0ab6e22feb9c72d4671af052a11d0badc4","cross_cats_sorted":["cs.AI"],"license":"http://creativecommons.org/licenses/by-nc-nd/4.0/","primary_cat":"cs.CL","submitted_at":"2024-10-09T19:03:52Z","title_canon_sha256":"518ca6e3587975b4a8e3d374eb5960e927901f852cf2ac4be736cd145b2b1c6a"},"schema_version":"1.0","source":{"id":"2410.07383","kind":"arxiv","version":1}},"canonical_sha256":"20b64a16802e43e30e9282fc78f290dd65678a4808e84f6ddc2ed0b256422aa9","receipt":{"algorithm":"ed25519","builder_version":"pith-number-builder-2026-05-17-v1","canonical_sha256":"20b64a16802e43e30e9282fc78f290dd65678a4808e84f6ddc2ed0b256422aa9","first_computed_at":"2026-07-05T09:18:35.444219Z","key_id":"pith-v1-2026-05","kind":"pith_receipt","last_reissued_at":"2026-07-05T09:18:35.444219Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","receipt_version":"0.3","signature_b64":"h+YGqRzIe6q+GVuLyftYAQ0lKbzvkG4wFG74z7koBXIeBL0WgQUbulwf1nw0DpoeeGPT1slShbjekSc5RZZqAA==","signature_status":"signed_v1","signed_at":"2026-07-05T09:18:35.444638Z","signed_message":"canonical_sha256_bytes"},"source_id":"2410.07383","source_kind":"arxiv","source_version":1}}},"equivocations":[],"invalid_events":[],"applied_event_ids":["sha256:fae5c54afc4f9e6b0f7689519f82c1d157447d789d4edbd366be31f1d9088adb","sha256:ac653d1bef5622c7eb9573dfc7a54d27a3910b2a0ca7447fdede1cf2eed52515"],"state_sha256":"499bfd806ff11b1d8f58fc9c781cb68c083828482aa2a3df5ffdc37deaf36733"},"bundle_signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"na8GDa65dJyB365luNICPl3ejQMrmMg8Jfd0pDYXbiR6twOiUoWCT/VMQxEx7rzRfsCetZGljtXM5LBFvgtRAw==","signed_message":"bundle_sha256_bytes","signed_at":"2026-07-10T03:32:05.093262Z","bundle_sha256":"3698553e9678fe91e29780fb06230ed0fb395c254656a252eddadf4c3ecd1494"}}