{"bundle_type":"pith_open_graph_bundle","bundle_version":"1.0","pith_number":"pith:2020:CSUTGYMODOYM3J3EEUQZH6OBUH","short_pith_number":"pith:CSUTGYMO","canonical_record":{"source":{"id":"2004.05529","kind":"arxiv","version":1},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.LG","submitted_at":"2020-04-12T02:57:28Z","cross_cats_sorted":["cs.CV","stat.ML"],"title_canon_sha256":"b68dbb1738dc7fb0b581964b73cde865e5271d6637b31f9beab59a1554be8626","abstract_canon_sha256":"31356bc5269606b014b3fa967084787e2b052cc238ed59c4543a403c54556809"},"schema_version":"1.0"},"canonical_sha256":"14a933618e1bb0cda764252193f9c1a1ea8b5ee5703c2cb4d933cd7de611242f","source":{"kind":"arxiv","id":"2004.05529","version":1},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2004.05529","created_at":"2026-07-05T00:54:38Z"},{"alias_kind":"arxiv_version","alias_value":"2004.05529v1","created_at":"2026-07-05T00:54:38Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2004.05529","created_at":"2026-07-05T00:54:38Z"},{"alias_kind":"pith_short_12","alias_value":"CSUTGYMODOYM","created_at":"2026-07-05T00:54:38Z"},{"alias_kind":"pith_short_16","alias_value":"CSUTGYMODOYM3J3E","created_at":"2026-07-05T00:54:38Z"},{"alias_kind":"pith_short_8","alias_value":"CSUTGYMO","created_at":"2026-07-05T00:54:38Z"}],"events":[{"event_type":"record_created","subject_pith_number":"pith:2020:CSUTGYMODOYM3J3EEUQZH6OBUH","target":"record","payload":{"canonical_record":{"source":{"id":"2004.05529","kind":"arxiv","version":1},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.LG","submitted_at":"2020-04-12T02:57:28Z","cross_cats_sorted":["cs.CV","stat.ML"],"title_canon_sha256":"b68dbb1738dc7fb0b581964b73cde865e5271d6637b31f9beab59a1554be8626","abstract_canon_sha256":"31356bc5269606b014b3fa967084787e2b052cc238ed59c4543a403c54556809"},"schema_version":"1.0"},"canonical_sha256":"14a933618e1bb0cda764252193f9c1a1ea8b5ee5703c2cb4d933cd7de611242f","receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T00:54:38.070785Z","signature_b64":"WzJkcs+pTey6y4ThUhyI/pMxt2Z8hhz9KiXkMUHibjM9aJFfL9PGmkw5AHicZKfJ+HjPfxY6O7ntp5GG4RaVDA==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"14a933618e1bb0cda764252193f9c1a1ea8b5ee5703c2cb4d933cd7de611242f","last_reissued_at":"2026-07-05T00:54:38.070343Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T00:54:38.070343Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"source_kind":"arxiv","source_id":"2004.05529","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":"2026-07-05T00:54:38Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"smC+N5529Ft630qiZJyHrnDc9PfpcsLl0B9RoS5a2kpGP7Klzq1UjuxuyxqEatKFuJl77b+fPmgSxcG2tbFFAw==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-08-09T11:54:43.822053Z"},"content_sha256":"b3c5db072da223f42b3b3d11df8b64aaddddbb791df40d1ca5df8fb33b23d3ab","schema_version":"1.0","event_id":"sha256:b3c5db072da223f42b3b3d11df8b64aaddddbb791df40d1ca5df8fb33b23d3ab"},{"event_type":"graph_snapshot","subject_pith_number":"pith:2020:CSUTGYMODOYM3J3EEUQZH6OBUH","target":"graph","payload":{"graph_snapshot":{"paper":{"title":"Gradients as Features for Deep Representation Learning","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.CV","stat.ML"],"primary_cat":"cs.LG","authors_text":"Fangzhou Mu, Yingyu Liang, Yin Li","submitted_at":"2020-04-12T02:57:28Z","abstract_excerpt":"We address the challenging problem of deep representation learning--the efficient adaption of a pre-trained deep network to different tasks. Specifically, we propose to explore gradient-based features. These features are gradients of the model parameters with respect to a task-specific loss given an input sample. Our key innovation is the design of a linear model that incorporates both gradient and activation of the pre-trained network. We show that our model provides a local linear approximation to an underlying deep model, and discuss important theoretical insights. Moreover, we present an e"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2004.05529","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/2004.05529/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":"2026-07-05T00:54:38Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"U77sgN2HtbKMjLp5Ts7V6n+IIcoNJL3Q/Pv/ClI7EgpJMOK97FOAohNASBRK2SJQBs0wDtDIYG6zS0oUKjbJAg==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-08-09T11:54:43.824247Z"},"content_sha256":"ffea8642f8df8cde12a48423298b8db4de6ba00a47580993907d8f740fa88837","schema_version":"1.0","event_id":"sha256:ffea8642f8df8cde12a48423298b8db4de6ba00a47580993907d8f740fa88837"}],"timestamp_proofs":[],"mirror_hints":[{"mirror_type":"https","name":"Pith Resolver","base_url":"https://pith.science","bundle_url":"https://pith.science/pith/CSUTGYMODOYM3J3EEUQZH6OBUH/bundle.json","state_url":"https://pith.science/pith/CSUTGYMODOYM3J3EEUQZH6OBUH/state.json","well_known_bundle_url":"https://pith.science/.well-known/pith/CSUTGYMODOYM3J3EEUQZH6OBUH/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":"2026-08-09T11:54:43Z","links":{"resolver":"https://pith.science/pith/CSUTGYMODOYM3J3EEUQZH6OBUH","bundle":"https://pith.science/pith/CSUTGYMODOYM3J3EEUQZH6OBUH/bundle.json","state":"https://pith.science/pith/CSUTGYMODOYM3J3EEUQZH6OBUH/state.json","well_known_bundle":"https://pith.science/.well-known/pith/CSUTGYMODOYM3J3EEUQZH6OBUH/bundle.json"},"state":{"state_type":"pith_open_graph_state","state_version":"1.0","pith_number":"pith:2020:CSUTGYMODOYM3J3EEUQZH6OBUH","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":"31356bc5269606b014b3fa967084787e2b052cc238ed59c4543a403c54556809","cross_cats_sorted":["cs.CV","stat.ML"],"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.LG","submitted_at":"2020-04-12T02:57:28Z","title_canon_sha256":"b68dbb1738dc7fb0b581964b73cde865e5271d6637b31f9beab59a1554be8626"},"schema_version":"1.0","source":{"id":"2004.05529","kind":"arxiv","version":1}},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2004.05529","created_at":"2026-07-05T00:54:38Z"},{"alias_kind":"arxiv_version","alias_value":"2004.05529v1","created_at":"2026-07-05T00:54:38Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2004.05529","created_at":"2026-07-05T00:54:38Z"},{"alias_kind":"pith_short_12","alias_value":"CSUTGYMODOYM","created_at":"2026-07-05T00:54:38Z"},{"alias_kind":"pith_short_16","alias_value":"CSUTGYMODOYM3J3E","created_at":"2026-07-05T00:54:38Z"},{"alias_kind":"pith_short_8","alias_value":"CSUTGYMO","created_at":"2026-07-05T00:54:38Z"}],"graph_snapshots":[{"event_id":"sha256:ffea8642f8df8cde12a48423298b8db4de6ba00a47580993907d8f740fa88837","target":"graph","created_at":"2026-07-05T00:54:38Z","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/2004.05529/integrity.json","findings":[],"snapshot_sha256":"c28c3603d3b5d939e8dc4c7e95fa8dfce3d595e45f758748cecf8e644a296938","summary":{"advisory":0,"by_detector":{},"critical":0,"informational":0}},"paper":{"abstract_excerpt":"We address the challenging problem of deep representation learning--the efficient adaption of a pre-trained deep network to different tasks. Specifically, we propose to explore gradient-based features. These features are gradients of the model parameters with respect to a task-specific loss given an input sample. Our key innovation is the design of a linear model that incorporates both gradient and activation of the pre-trained network. We show that our model provides a local linear approximation to an underlying deep model, and discuss important theoretical insights. Moreover, we present an e","authors_text":"Fangzhou Mu, Yingyu Liang, Yin Li","cross_cats":["cs.CV","stat.ML"],"headline":"","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.LG","submitted_at":"2020-04-12T02:57:28Z","title":"Gradients as Features for Deep Representation Learning"},"references":{"count":0,"internal_anchors":0,"resolved_work":0,"sample":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2004.05529","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:b3c5db072da223f42b3b3d11df8b64aaddddbb791df40d1ca5df8fb33b23d3ab","target":"record","created_at":"2026-07-05T00:54:38Z","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":"31356bc5269606b014b3fa967084787e2b052cc238ed59c4543a403c54556809","cross_cats_sorted":["cs.CV","stat.ML"],"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.LG","submitted_at":"2020-04-12T02:57:28Z","title_canon_sha256":"b68dbb1738dc7fb0b581964b73cde865e5271d6637b31f9beab59a1554be8626"},"schema_version":"1.0","source":{"id":"2004.05529","kind":"arxiv","version":1}},"canonical_sha256":"14a933618e1bb0cda764252193f9c1a1ea8b5ee5703c2cb4d933cd7de611242f","receipt":{"algorithm":"ed25519","builder_version":"pith-number-builder-2026-05-17-v1","canonical_sha256":"14a933618e1bb0cda764252193f9c1a1ea8b5ee5703c2cb4d933cd7de611242f","first_computed_at":"2026-07-05T00:54:38.070343Z","key_id":"pith-v1-2026-05","kind":"pith_receipt","last_reissued_at":"2026-07-05T00:54:38.070343Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","receipt_version":"0.3","signature_b64":"WzJkcs+pTey6y4ThUhyI/pMxt2Z8hhz9KiXkMUHibjM9aJFfL9PGmkw5AHicZKfJ+HjPfxY6O7ntp5GG4RaVDA==","signature_status":"signed_v1","signed_at":"2026-07-05T00:54:38.070785Z","signed_message":"canonical_sha256_bytes"},"source_id":"2004.05529","source_kind":"arxiv","source_version":1}}},"equivocations":[],"invalid_events":[],"applied_event_ids":["sha256:b3c5db072da223f42b3b3d11df8b64aaddddbb791df40d1ca5df8fb33b23d3ab","sha256:ffea8642f8df8cde12a48423298b8db4de6ba00a47580993907d8f740fa88837"],"state_sha256":"2990a8ffe530105a09f5d82da266aedc6a68bf2d3b76af9c60faf8f01eebac86"},"bundle_signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"L2LOP7clURrTwZjypKUgNfFh2LKIz5pbBWfno0HoRnbskCeB3vTl3TTXK1N7V3aXX52ikiZy9QnVhHnMR+w+CA==","signed_message":"bundle_sha256_bytes","signed_at":"2026-08-09T11:54:43.832285Z","bundle_sha256":"dfed360ef4d32f31fb9e1d442f627675b94392068713f3d30be85f2133918cda"}}