{"bundle_type":"pith_open_graph_bundle","bundle_version":"1.0","pith_number":"pith:2024:36GUYLOSUHKR3WU4NGK7RVINMB","short_pith_number":"pith:36GUYLOS","canonical_record":{"source":{"id":"2403.06213","kind":"arxiv","version":1},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CV","submitted_at":"2024-03-10T13:26:24Z","cross_cats_sorted":["cs.AI"],"title_canon_sha256":"49b8dabbf2cf49748d20d06e994cc24edb70a12abae95936576fd73724139bdf","abstract_canon_sha256":"8c2e250ea177e1bb65d83175c822bdd4b6dc4b16d3f220d4110d9897dd1b5148"},"schema_version":"1.0"},"canonical_sha256":"df8d4c2dd2a1d51dda9c6995f8d50d6063e6311bb0d1f3d303c0329a46c22928","source":{"kind":"arxiv","id":"2403.06213","version":1},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2403.06213","created_at":"2026-07-05T07:54:20Z"},{"alias_kind":"arxiv_version","alias_value":"2403.06213v1","created_at":"2026-07-05T07:54:20Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2403.06213","created_at":"2026-07-05T07:54:20Z"},{"alias_kind":"pith_short_12","alias_value":"36GUYLOSUHKR","created_at":"2026-07-05T07:54:20Z"},{"alias_kind":"pith_short_16","alias_value":"36GUYLOSUHKR3WU4","created_at":"2026-07-05T07:54:20Z"},{"alias_kind":"pith_short_8","alias_value":"36GUYLOS","created_at":"2026-07-05T07:54:20Z"}],"events":[{"event_type":"record_created","subject_pith_number":"pith:2024:36GUYLOSUHKR3WU4NGK7RVINMB","target":"record","payload":{"canonical_record":{"source":{"id":"2403.06213","kind":"arxiv","version":1},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CV","submitted_at":"2024-03-10T13:26:24Z","cross_cats_sorted":["cs.AI"],"title_canon_sha256":"49b8dabbf2cf49748d20d06e994cc24edb70a12abae95936576fd73724139bdf","abstract_canon_sha256":"8c2e250ea177e1bb65d83175c822bdd4b6dc4b16d3f220d4110d9897dd1b5148"},"schema_version":"1.0"},"canonical_sha256":"df8d4c2dd2a1d51dda9c6995f8d50d6063e6311bb0d1f3d303c0329a46c22928","receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T07:54:20.137878Z","signature_b64":"gBq9WSpI1z+m40dQ9XE2+qvSleeqTmuD9lku5ItW0nS5y48sGFisPqae++cz0wIkuWkYbP0lFTUB41+ZnclYBg==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"df8d4c2dd2a1d51dda9c6995f8d50d6063e6311bb0d1f3d303c0329a46c22928","last_reissued_at":"2026-07-05T07:54:20.137463Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T07:54:20.137463Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"source_kind":"arxiv","source_id":"2403.06213","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":"2026-07-05T07:54:20Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"21BHC19TXIPTLn7ggv2lI2nclxcUGGSjlONDbI73vT63fdknjoZs+knpnYLW2xWTHB11S5VqEEtLMFoQeEQCDA==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-08-09T22:31:23.852106Z"},"content_sha256":"e249f7044ed0b014a8956433e0dbf087c308d3fe86a3d642dfd4939467eecc6e","schema_version":"1.0","event_id":"sha256:e249f7044ed0b014a8956433e0dbf087c308d3fe86a3d642dfd4939467eecc6e"},{"event_type":"graph_snapshot","subject_pith_number":"pith:2024:36GUYLOSUHKR3WU4NGK7RVINMB","target":"graph","payload":{"graph_snapshot":{"paper":{"title":"$V_kD:$ Improving Knowledge Distillation using Orthogonal Projections","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.AI"],"primary_cat":"cs.CV","authors_text":"Ismail Elezi, Jiankang Deng, Roy Miles","submitted_at":"2024-03-10T13:26:24Z","abstract_excerpt":"Knowledge distillation is an effective method for training small and efficient deep learning models. However, the efficacy of a single method can degenerate when transferring to other tasks, modalities, or even other architectures. To address this limitation, we propose a novel constrained feature distillation method. This method is derived from a small set of core principles, which results in two emerging components: an orthogonal projection and a task-specific normalisation. Equipped with both of these components, our transformer models can outperform all previous methods on ImageNet and rea"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2403.06213","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/2403.06213/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":"2026-07-05T07:54:20Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"3jXEBEFMVEQIJPx6VpSidJlHKJkxuluvYLUG0ithEc/n/Kkvi9w4q+aEkv6bBqn7GZKMqsJkDIjltedklN/eAQ==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-08-09T22:31:23.852587Z"},"content_sha256":"52ec14a67f829b89d9450cfd78502019920e79010a9e95839e2a10206b84e8d6","schema_version":"1.0","event_id":"sha256:52ec14a67f829b89d9450cfd78502019920e79010a9e95839e2a10206b84e8d6"}],"timestamp_proofs":[],"mirror_hints":[{"mirror_type":"https","name":"Pith Resolver","base_url":"https://pith.science","bundle_url":"https://pith.science/pith/36GUYLOSUHKR3WU4NGK7RVINMB/bundle.json","state_url":"https://pith.science/pith/36GUYLOSUHKR3WU4NGK7RVINMB/state.json","well_known_bundle_url":"https://pith.science/.well-known/pith/36GUYLOSUHKR3WU4NGK7RVINMB/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":"2026-08-09T22:31:23Z","links":{"resolver":"https://pith.science/pith/36GUYLOSUHKR3WU4NGK7RVINMB","bundle":"https://pith.science/pith/36GUYLOSUHKR3WU4NGK7RVINMB/bundle.json","state":"https://pith.science/pith/36GUYLOSUHKR3WU4NGK7RVINMB/state.json","well_known_bundle":"https://pith.science/.well-known/pith/36GUYLOSUHKR3WU4NGK7RVINMB/bundle.json"},"state":{"state_type":"pith_open_graph_state","state_version":"1.0","pith_number":"pith:2024:36GUYLOSUHKR3WU4NGK7RVINMB","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":"8c2e250ea177e1bb65d83175c822bdd4b6dc4b16d3f220d4110d9897dd1b5148","cross_cats_sorted":["cs.AI"],"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CV","submitted_at":"2024-03-10T13:26:24Z","title_canon_sha256":"49b8dabbf2cf49748d20d06e994cc24edb70a12abae95936576fd73724139bdf"},"schema_version":"1.0","source":{"id":"2403.06213","kind":"arxiv","version":1}},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2403.06213","created_at":"2026-07-05T07:54:20Z"},{"alias_kind":"arxiv_version","alias_value":"2403.06213v1","created_at":"2026-07-05T07:54:20Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2403.06213","created_at":"2026-07-05T07:54:20Z"},{"alias_kind":"pith_short_12","alias_value":"36GUYLOSUHKR","created_at":"2026-07-05T07:54:20Z"},{"alias_kind":"pith_short_16","alias_value":"36GUYLOSUHKR3WU4","created_at":"2026-07-05T07:54:20Z"},{"alias_kind":"pith_short_8","alias_value":"36GUYLOS","created_at":"2026-07-05T07:54:20Z"}],"graph_snapshots":[{"event_id":"sha256:52ec14a67f829b89d9450cfd78502019920e79010a9e95839e2a10206b84e8d6","target":"graph","created_at":"2026-07-05T07:54:20Z","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/2403.06213/integrity.json","findings":[],"snapshot_sha256":"c28c3603d3b5d939e8dc4c7e95fa8dfce3d595e45f758748cecf8e644a296938","summary":{"advisory":0,"by_detector":{},"critical":0,"informational":0}},"paper":{"abstract_excerpt":"Knowledge distillation is an effective method for training small and efficient deep learning models. However, the efficacy of a single method can degenerate when transferring to other tasks, modalities, or even other architectures. To address this limitation, we propose a novel constrained feature distillation method. This method is derived from a small set of core principles, which results in two emerging components: an orthogonal projection and a task-specific normalisation. Equipped with both of these components, our transformer models can outperform all previous methods on ImageNet and rea","authors_text":"Ismail Elezi, Jiankang Deng, Roy Miles","cross_cats":["cs.AI"],"headline":"","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CV","submitted_at":"2024-03-10T13:26:24Z","title":"$V_kD:$ Improving Knowledge Distillation using Orthogonal Projections"},"references":{"count":0,"internal_anchors":0,"resolved_work":0,"sample":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2403.06213","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:e249f7044ed0b014a8956433e0dbf087c308d3fe86a3d642dfd4939467eecc6e","target":"record","created_at":"2026-07-05T07:54:20Z","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":"8c2e250ea177e1bb65d83175c822bdd4b6dc4b16d3f220d4110d9897dd1b5148","cross_cats_sorted":["cs.AI"],"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CV","submitted_at":"2024-03-10T13:26:24Z","title_canon_sha256":"49b8dabbf2cf49748d20d06e994cc24edb70a12abae95936576fd73724139bdf"},"schema_version":"1.0","source":{"id":"2403.06213","kind":"arxiv","version":1}},"canonical_sha256":"df8d4c2dd2a1d51dda9c6995f8d50d6063e6311bb0d1f3d303c0329a46c22928","receipt":{"algorithm":"ed25519","builder_version":"pith-number-builder-2026-05-17-v1","canonical_sha256":"df8d4c2dd2a1d51dda9c6995f8d50d6063e6311bb0d1f3d303c0329a46c22928","first_computed_at":"2026-07-05T07:54:20.137463Z","key_id":"pith-v1-2026-05","kind":"pith_receipt","last_reissued_at":"2026-07-05T07:54:20.137463Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","receipt_version":"0.3","signature_b64":"gBq9WSpI1z+m40dQ9XE2+qvSleeqTmuD9lku5ItW0nS5y48sGFisPqae++cz0wIkuWkYbP0lFTUB41+ZnclYBg==","signature_status":"signed_v1","signed_at":"2026-07-05T07:54:20.137878Z","signed_message":"canonical_sha256_bytes"},"source_id":"2403.06213","source_kind":"arxiv","source_version":1}}},"equivocations":[],"invalid_events":[],"applied_event_ids":["sha256:e249f7044ed0b014a8956433e0dbf087c308d3fe86a3d642dfd4939467eecc6e","sha256:52ec14a67f829b89d9450cfd78502019920e79010a9e95839e2a10206b84e8d6"],"state_sha256":"c6edba2d2b1a62db0284749bf3d5982bd6efdce899aa12f258d96ce6dfe2ef27"},"bundle_signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"IhH/EcjKNjm2c3U6GdQIn+LRu/nquEQO6GR18LEtLzgkLnfHJLKZNW8p/KHDDiFVs/uk3O6WbSzgUDPDE8JtDw==","signed_message":"bundle_sha256_bytes","signed_at":"2026-08-09T22:31:23.856054Z","bundle_sha256":"af05d5feac2354a5c9b1bebd95d1f6e71ed8deb851e28d90b338c107cb39053a"}}