{"bundle_type":"pith_open_graph_bundle","bundle_version":"1.0","pith_number":"pith:2024:S4CGGMDHRPJPYJYEDIPSHWQQR4","short_pith_number":"pith:S4CGGMDH","canonical_record":{"source":{"id":"2407.20229","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CV","submitted_at":"2024-07-29T17:59:21Z","cross_cats_sorted":[],"title_canon_sha256":"6f654c70e77ef71e21e32bf247bd5514e6570221f8a8d1adbaf49c6f02653686","abstract_canon_sha256":"5b991b8290d60ec451090136b989657fccbe5117fe89da8e360945c803303d64"},"schema_version":"1.0"},"canonical_sha256":"97046330678bd2fc27041a1f23da108f0a60a79b959d2f9bafe90e40d4acd2f4","source":{"kind":"arxiv","id":"2407.20229","version":1},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2407.20229","created_at":"2026-07-05T08:49:50Z"},{"alias_kind":"arxiv_version","alias_value":"2407.20229v1","created_at":"2026-07-05T08:49:50Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2407.20229","created_at":"2026-07-05T08:49:50Z"},{"alias_kind":"pith_short_12","alias_value":"S4CGGMDHRPJP","created_at":"2026-07-05T08:49:50Z"},{"alias_kind":"pith_short_16","alias_value":"S4CGGMDHRPJPYJYE","created_at":"2026-07-05T08:49:50Z"},{"alias_kind":"pith_short_8","alias_value":"S4CGGMDH","created_at":"2026-07-05T08:49:50Z"}],"events":[{"event_type":"record_created","subject_pith_number":"pith:2024:S4CGGMDHRPJPYJYEDIPSHWQQR4","target":"record","payload":{"canonical_record":{"source":{"id":"2407.20229","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CV","submitted_at":"2024-07-29T17:59:21Z","cross_cats_sorted":[],"title_canon_sha256":"6f654c70e77ef71e21e32bf247bd5514e6570221f8a8d1adbaf49c6f02653686","abstract_canon_sha256":"5b991b8290d60ec451090136b989657fccbe5117fe89da8e360945c803303d64"},"schema_version":"1.0"},"canonical_sha256":"97046330678bd2fc27041a1f23da108f0a60a79b959d2f9bafe90e40d4acd2f4","receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T08:49:50.162284Z","signature_b64":"MnkHH1g6wNWdONDffnGu9XSqKOHlxE8e4B3risatuNoUq/op/mvxGnO+Rm8dBrY4BoGuYYGlyxEHHMZcyAMDAg==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"97046330678bd2fc27041a1f23da108f0a60a79b959d2f9bafe90e40d4acd2f4","last_reissued_at":"2026-07-05T08:49:50.161873Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T08:49:50.161873Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"source_kind":"arxiv","source_id":"2407.20229","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":"2026-07-05T08:49:50Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"f+4QTQkzkAOPT2zSNdMssCFKb7uf6+xeulsrn6JxOm9CxW8b+C4IlANqXpkKOmNkYUm3VhFV7kBZDzsolkFmBw==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-07-07T13:09:43.396343Z"},"content_sha256":"a7746d33620dfe2defe6d44bc05fe20ca9144366298c81177fab9afd4cd1507c","schema_version":"1.0","event_id":"sha256:a7746d33620dfe2defe6d44bc05fe20ca9144366298c81177fab9afd4cd1507c"},{"event_type":"graph_snapshot","subject_pith_number":"pith:2024:S4CGGMDHRPJPYJYEDIPSHWQQR4","target":"graph","payload":{"graph_snapshot":{"paper":{"title":"Improving 2D Feature Representations by 3D-Aware Fine-Tuning","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":[],"primary_cat":"cs.CV","authors_text":"Anurag Das, Francis Engelmann, Jan Eric Lenssen, Siyu Tang, Yuanwen Yue","submitted_at":"2024-07-29T17:59:21Z","abstract_excerpt":"Current visual foundation models are trained purely on unstructured 2D data, limiting their understanding of 3D structure of objects and scenes. In this work, we show that fine-tuning on 3D-aware data improves the quality of emerging semantic features. We design a method to lift semantic 2D features into an efficient 3D Gaussian representation, which allows us to re-render them for arbitrary views. Using the rendered 3D-aware features, we design a fine-tuning strategy to transfer such 3D awareness into a 2D foundation model. We demonstrate that models fine-tuned in that way produce features th"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2407.20229","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/2407.20229/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":"2026-07-05T08:49:50Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"g3rPC6hrga6TOEqN51W+Isc0x1n0EYki9hAZLqaFc+QARJamuByYfdELIO6A7D+AYntD4G3PgqvgnXYVFB/XBA==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-07-07T13:09:43.396718Z"},"content_sha256":"a36e4f9e798ace5618118dc41ff2cdf80215cbf3dbcbeaea31acc078d29f7e28","schema_version":"1.0","event_id":"sha256:a36e4f9e798ace5618118dc41ff2cdf80215cbf3dbcbeaea31acc078d29f7e28"}],"timestamp_proofs":[],"mirror_hints":[{"mirror_type":"https","name":"Pith Resolver","base_url":"https://pith.science","bundle_url":"https://pith.science/pith/S4CGGMDHRPJPYJYEDIPSHWQQR4/bundle.json","state_url":"https://pith.science/pith/S4CGGMDHRPJPYJYEDIPSHWQQR4/state.json","well_known_bundle_url":"https://pith.science/.well-known/pith/S4CGGMDHRPJPYJYEDIPSHWQQR4/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":"2026-07-07T13:09:43Z","links":{"resolver":"https://pith.science/pith/S4CGGMDHRPJPYJYEDIPSHWQQR4","bundle":"https://pith.science/pith/S4CGGMDHRPJPYJYEDIPSHWQQR4/bundle.json","state":"https://pith.science/pith/S4CGGMDHRPJPYJYEDIPSHWQQR4/state.json","well_known_bundle":"https://pith.science/.well-known/pith/S4CGGMDHRPJPYJYEDIPSHWQQR4/bundle.json"},"state":{"state_type":"pith_open_graph_state","state_version":"1.0","pith_number":"pith:2024:S4CGGMDHRPJPYJYEDIPSHWQQR4","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":"5b991b8290d60ec451090136b989657fccbe5117fe89da8e360945c803303d64","cross_cats_sorted":[],"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CV","submitted_at":"2024-07-29T17:59:21Z","title_canon_sha256":"6f654c70e77ef71e21e32bf247bd5514e6570221f8a8d1adbaf49c6f02653686"},"schema_version":"1.0","source":{"id":"2407.20229","kind":"arxiv","version":1}},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2407.20229","created_at":"2026-07-05T08:49:50Z"},{"alias_kind":"arxiv_version","alias_value":"2407.20229v1","created_at":"2026-07-05T08:49:50Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2407.20229","created_at":"2026-07-05T08:49:50Z"},{"alias_kind":"pith_short_12","alias_value":"S4CGGMDHRPJP","created_at":"2026-07-05T08:49:50Z"},{"alias_kind":"pith_short_16","alias_value":"S4CGGMDHRPJPYJYE","created_at":"2026-07-05T08:49:50Z"},{"alias_kind":"pith_short_8","alias_value":"S4CGGMDH","created_at":"2026-07-05T08:49:50Z"}],"graph_snapshots":[{"event_id":"sha256:a36e4f9e798ace5618118dc41ff2cdf80215cbf3dbcbeaea31acc078d29f7e28","target":"graph","created_at":"2026-07-05T08:49:50Z","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/2407.20229/integrity.json","findings":[],"snapshot_sha256":"c28c3603d3b5d939e8dc4c7e95fa8dfce3d595e45f758748cecf8e644a296938","summary":{"advisory":0,"by_detector":{},"critical":0,"informational":0}},"paper":{"abstract_excerpt":"Current visual foundation models are trained purely on unstructured 2D data, limiting their understanding of 3D structure of objects and scenes. In this work, we show that fine-tuning on 3D-aware data improves the quality of emerging semantic features. We design a method to lift semantic 2D features into an efficient 3D Gaussian representation, which allows us to re-render them for arbitrary views. Using the rendered 3D-aware features, we design a fine-tuning strategy to transfer such 3D awareness into a 2D foundation model. We demonstrate that models fine-tuned in that way produce features th","authors_text":"Anurag Das, Francis Engelmann, Jan Eric Lenssen, Siyu Tang, Yuanwen Yue","cross_cats":[],"headline":"","license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CV","submitted_at":"2024-07-29T17:59:21Z","title":"Improving 2D Feature Representations by 3D-Aware Fine-Tuning"},"references":{"count":0,"internal_anchors":0,"resolved_work":0,"sample":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2407.20229","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:a7746d33620dfe2defe6d44bc05fe20ca9144366298c81177fab9afd4cd1507c","target":"record","created_at":"2026-07-05T08:49:50Z","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":"5b991b8290d60ec451090136b989657fccbe5117fe89da8e360945c803303d64","cross_cats_sorted":[],"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CV","submitted_at":"2024-07-29T17:59:21Z","title_canon_sha256":"6f654c70e77ef71e21e32bf247bd5514e6570221f8a8d1adbaf49c6f02653686"},"schema_version":"1.0","source":{"id":"2407.20229","kind":"arxiv","version":1}},"canonical_sha256":"97046330678bd2fc27041a1f23da108f0a60a79b959d2f9bafe90e40d4acd2f4","receipt":{"algorithm":"ed25519","builder_version":"pith-number-builder-2026-05-17-v1","canonical_sha256":"97046330678bd2fc27041a1f23da108f0a60a79b959d2f9bafe90e40d4acd2f4","first_computed_at":"2026-07-05T08:49:50.161873Z","key_id":"pith-v1-2026-05","kind":"pith_receipt","last_reissued_at":"2026-07-05T08:49:50.161873Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","receipt_version":"0.3","signature_b64":"MnkHH1g6wNWdONDffnGu9XSqKOHlxE8e4B3risatuNoUq/op/mvxGnO+Rm8dBrY4BoGuYYGlyxEHHMZcyAMDAg==","signature_status":"signed_v1","signed_at":"2026-07-05T08:49:50.162284Z","signed_message":"canonical_sha256_bytes"},"source_id":"2407.20229","source_kind":"arxiv","source_version":1}}},"equivocations":[],"invalid_events":[],"applied_event_ids":["sha256:a7746d33620dfe2defe6d44bc05fe20ca9144366298c81177fab9afd4cd1507c","sha256:a36e4f9e798ace5618118dc41ff2cdf80215cbf3dbcbeaea31acc078d29f7e28"],"state_sha256":"e0f2603d9354404fc30da12123a382b0812d3ac2940b5bc116f980a6a0ec90e4"},"bundle_signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"tL+ukaumaTV1tphb3uXU7UhbA156qB9dfJP84e72leJO8qJTjzk1a/La7Pao9Rz33rcb3+L7E/SUIDQ9pq5jBg==","signed_message":"bundle_sha256_bytes","signed_at":"2026-07-07T13:09:43.399149Z","bundle_sha256":"697be616dd420f8b39257953ea0c41a3c1ba770f45f22bf20dd7d0fd8a9dca6e"}}