{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2022:IVKMT2P6DBXGTQ5L7IVFLXFBDS","short_pith_number":"pith:IVKMT2P6","schema_version":"1.0","canonical_sha256":"4554c9e9fe186e69c3abfa2a55dca11c9398496c3c8887c35ac043391a62a341","source":{"kind":"arxiv","id":"2209.15430","version":2},"attestation_state":"computed","paper":{"title":"Relative representations enable zero-shot latent space communication","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.AI"],"primary_cat":"cs.LG","authors_text":"Antonio Norelli, Emanuele Rodol\\`a, Francesco Locatello, Luca Moschella, Marco Fumero, Valentino Maiorca","submitted_at":"2022-09-30T12:37:03Z","abstract_excerpt":"Neural networks embed the geometric structure of a data manifold lying in a high-dimensional space into latent representations. Ideally, the distribution of the data points in the latent space should depend only on the task, the data, the loss, and other architecture-specific constraints. However, factors such as the random weights initialization, training hyperparameters, or other sources of randomness in the training phase may induce incoherent latent spaces that hinder any form of reuse. Nevertheless, we empirically observe that, under the same data and modeling choices, the angles between "},"verification_status":{"content_addressed":true,"pith_receipt":true,"author_attested":false,"weak_author_claims":0,"strong_author_claims":0,"externally_anchored":false,"storage_verified":false,"citation_signatures":0,"replication_records":0,"graph_snapshot":true,"references_resolved":false,"formal_links_present":false},"canonical_record":{"source":{"id":"2209.15430","kind":"arxiv","version":2},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.LG","submitted_at":"2022-09-30T12:37:03Z","cross_cats_sorted":["cs.AI"],"title_canon_sha256":"e3b0ff102cc417b040ecf94f0fac7898ff4dd42ee1061acaeb1543c9e7cce467","abstract_canon_sha256":"a671bf9747d8412fcb2b8051e85c3b36649adb01c4adb8b1fdaca28b3025c411"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T05:48:39.525256Z","signature_b64":"2H/TXUeEUKgj/ITchmrhj+VeLGAAubs4Ekph/FZxR4voISfoJlBgb4/N35+gdH/PcyaNO0JAchFxVImSnpD+AA==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"4554c9e9fe186e69c3abfa2a55dca11c9398496c3c8887c35ac043391a62a341","last_reissued_at":"2026-07-05T05:48:39.524776Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T05:48:39.524776Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Relative representations enable zero-shot latent space communication","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.AI"],"primary_cat":"cs.LG","authors_text":"Antonio Norelli, Emanuele Rodol\\`a, Francesco Locatello, Luca Moschella, Marco Fumero, Valentino Maiorca","submitted_at":"2022-09-30T12:37:03Z","abstract_excerpt":"Neural networks embed the geometric structure of a data manifold lying in a high-dimensional space into latent representations. Ideally, the distribution of the data points in the latent space should depend only on the task, the data, the loss, and other architecture-specific constraints. However, factors such as the random weights initialization, training hyperparameters, or other sources of randomness in the training phase may induce incoherent latent spaces that hinder any form of reuse. Nevertheless, we empirically observe that, under the same data and modeling choices, the angles between "},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2209.15430","kind":"arxiv","version":2},"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/2209.15430/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"},"aliases":[{"alias_kind":"arxiv","alias_value":"2209.15430","created_at":"2026-07-05T05:48:39.524844+00:00"},{"alias_kind":"arxiv_version","alias_value":"2209.15430v2","created_at":"2026-07-05T05:48:39.524844+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2209.15430","created_at":"2026-07-05T05:48:39.524844+00:00"},{"alias_kind":"pith_short_12","alias_value":"IVKMT2P6DBXG","created_at":"2026-07-05T05:48:39.524844+00:00"},{"alias_kind":"pith_short_16","alias_value":"IVKMT2P6DBXGTQ5L","created_at":"2026-07-05T05:48:39.524844+00:00"},{"alias_kind":"pith_short_8","alias_value":"IVKMT2P6","created_at":"2026-07-05T05:48:39.524844+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":10,"internal_anchor_count":0,"sample":[{"citing_arxiv_id":"2606.21509","citing_title":"A Stitch in Time Saves Nine: Preserving Policy Compatibility Under Perception Updates in End-to-End Autonomous Driving","ref_index":33,"is_internal_anchor":false},{"citing_arxiv_id":"2606.12005","citing_title":"Game-Theoretic Latent Space Alignment for Multi-user Semantic MIMO Communications","ref_index":17,"is_internal_anchor":false},{"citing_arxiv_id":"2606.06021","citing_title":"OPRD: On-Policy Representation Distillation","ref_index":27,"is_internal_anchor":false},{"citing_arxiv_id":"2605.30596","citing_title":"Improving Relative Representations with Learned Anchors and Whitened Inner Products","ref_index":1,"is_internal_anchor":false},{"citing_arxiv_id":"2502.18036","citing_title":"Harnessing Multiple Large Language Models: A Survey on LLM Ensemble","ref_index":35,"is_internal_anchor":false},{"citing_arxiv_id":"2511.06516","citing_title":"You Had One Job: Per-Task Quantization Using LLMs' Hidden Representations","ref_index":43,"is_internal_anchor":false},{"citing_arxiv_id":"2309.16797","citing_title":"Promptbreeder: Self-Referential Self-Improvement Via Prompt Evolution","ref_index":237,"is_internal_anchor":false},{"citing_arxiv_id":"2405.07987","citing_title":"The Platonic Representation Hypothesis","ref_index":288,"is_internal_anchor":false},{"citing_arxiv_id":"2311.05232","citing_title":"A Survey on Hallucination in Large Language Models: Principles, Taxonomy, Challenges, and Open Questions","ref_index":230,"is_internal_anchor":false},{"citing_arxiv_id":"2604.21691","citing_title":"There Will Be a Scientific Theory of Deep Learning","ref_index":6,"is_internal_anchor":false}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/IVKMT2P6DBXGTQ5L7IVFLXFBDS","json":"https://pith.science/pith/IVKMT2P6DBXGTQ5L7IVFLXFBDS.json","graph_json":"https://pith.science/api/pith-number/IVKMT2P6DBXGTQ5L7IVFLXFBDS/graph.json","events_json":"https://pith.science/api/pith-number/IVKMT2P6DBXGTQ5L7IVFLXFBDS/events.json","paper":"https://pith.science/paper/IVKMT2P6"},"agent_actions":{"view_html":"https://pith.science/pith/IVKMT2P6DBXGTQ5L7IVFLXFBDS","download_json":"https://pith.science/pith/IVKMT2P6DBXGTQ5L7IVFLXFBDS.json","view_paper":"https://pith.science/paper/IVKMT2P6","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2209.15430&json=true","fetch_graph":"https://pith.science/api/pith-number/IVKMT2P6DBXGTQ5L7IVFLXFBDS/graph.json","fetch_events":"https://pith.science/api/pith-number/IVKMT2P6DBXGTQ5L7IVFLXFBDS/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/IVKMT2P6DBXGTQ5L7IVFLXFBDS/action/timestamp_anchor","attest_storage":"https://pith.science/pith/IVKMT2P6DBXGTQ5L7IVFLXFBDS/action/storage_attestation","attest_author":"https://pith.science/pith/IVKMT2P6DBXGTQ5L7IVFLXFBDS/action/author_attestation","sign_citation":"https://pith.science/pith/IVKMT2P6DBXGTQ5L7IVFLXFBDS/action/citation_signature","submit_replication":"https://pith.science/pith/IVKMT2P6DBXGTQ5L7IVFLXFBDS/action/replication_record"}},"created_at":"2026-07-05T05:48:39.524844+00:00","updated_at":"2026-07-05T05:48:39.524844+00:00"}