{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2021:L7A2XEMS4GGY42BN5WH7AMKHZD","short_pith_number":"pith:L7A2XEMS","schema_version":"1.0","canonical_sha256":"5fc1ab9192e18d8e682ded8ff03147c8d4a60afefc72ece49dc9da4916d31906","source":{"kind":"arxiv","id":"2107.02561","version":3},"attestation_state":"computed","paper":{"title":"Rethinking Positional Encoding","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.CV"],"primary_cat":"cs.LG","authors_text":"Jianqiao Zheng, Sameera Ramasinghe, Simon Lucey","submitted_at":"2021-07-06T12:04:04Z","abstract_excerpt":"It is well noted that coordinate based MLPs benefit -- in terms of preserving high-frequency information -- through the encoding of coordinate positions as an array of Fourier features. Hitherto, the rationale for the effectiveness of these positional encodings has been solely studied through a Fourier lens. In this paper, we strive to broaden this understanding by showing that alternative non-Fourier embedding functions can indeed be used for positional encoding. Moreover, we show that their performance is entirely determined by a trade-off between the stable rank of the embedded matrix and t"},"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":"2107.02561","kind":"arxiv","version":3},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.LG","submitted_at":"2021-07-06T12:04:04Z","cross_cats_sorted":["cs.CV"],"title_canon_sha256":"0d4493e81fae85bafe31a8028ba833e5b93176eb29e4b25ba72a37fd6e8735da","abstract_canon_sha256":"19f49e200284e32d5e2afb17c58ceddc7188d569c3847d4d089aac351ef0b3fa"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T03:21:54.156616Z","signature_b64":"4HNXXlFUEOkBLKZzV9uumMcabFgJ5fzXOutY9Fmn03ktFZ9QYoGA1nRvHGjBB2vAwBn0A+tRQnlIAn9l9GPQAA==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"5fc1ab9192e18d8e682ded8ff03147c8d4a60afefc72ece49dc9da4916d31906","last_reissued_at":"2026-07-05T03:21:54.156083Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T03:21:54.156083Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Rethinking Positional Encoding","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.CV"],"primary_cat":"cs.LG","authors_text":"Jianqiao Zheng, Sameera Ramasinghe, Simon Lucey","submitted_at":"2021-07-06T12:04:04Z","abstract_excerpt":"It is well noted that coordinate based MLPs benefit -- in terms of preserving high-frequency information -- through the encoding of coordinate positions as an array of Fourier features. Hitherto, the rationale for the effectiveness of these positional encodings has been solely studied through a Fourier lens. In this paper, we strive to broaden this understanding by showing that alternative non-Fourier embedding functions can indeed be used for positional encoding. Moreover, we show that their performance is entirely determined by a trade-off between the stable rank of the embedded matrix and t"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2107.02561","kind":"arxiv","version":3},"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/2107.02561/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":"2107.02561","created_at":"2026-07-05T03:21:54.156157+00:00"},{"alias_kind":"arxiv_version","alias_value":"2107.02561v3","created_at":"2026-07-05T03:21:54.156157+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2107.02561","created_at":"2026-07-05T03:21:54.156157+00:00"},{"alias_kind":"pith_short_12","alias_value":"L7A2XEMS4GGY","created_at":"2026-07-05T03:21:54.156157+00:00"},{"alias_kind":"pith_short_16","alias_value":"L7A2XEMS4GGY42BN","created_at":"2026-07-05T03:21:54.156157+00:00"},{"alias_kind":"pith_short_8","alias_value":"L7A2XEMS","created_at":"2026-07-05T03:21:54.156157+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":3,"internal_anchor_count":0,"sample":[{"citing_arxiv_id":"2606.10756","citing_title":"DD-INR: Dynamics-Driven Implicit Neural Representation for Accelerated Whole-Brain Functional MRI Reconstruction","ref_index":29,"is_internal_anchor":false},{"citing_arxiv_id":"2605.15216","citing_title":"Hardware-Software Co-Design of Scalable, Energy-Efficient Analog Recurrent Computations","ref_index":103,"is_internal_anchor":false},{"citing_arxiv_id":"2605.15216","citing_title":"Hardware-Software Co-Design of Scalable, Energy-Efficient Analog Recurrent Computations","ref_index":103,"is_internal_anchor":false}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/L7A2XEMS4GGY42BN5WH7AMKHZD","json":"https://pith.science/pith/L7A2XEMS4GGY42BN5WH7AMKHZD.json","graph_json":"https://pith.science/api/pith-number/L7A2XEMS4GGY42BN5WH7AMKHZD/graph.json","events_json":"https://pith.science/api/pith-number/L7A2XEMS4GGY42BN5WH7AMKHZD/events.json","paper":"https://pith.science/paper/L7A2XEMS"},"agent_actions":{"view_html":"https://pith.science/pith/L7A2XEMS4GGY42BN5WH7AMKHZD","download_json":"https://pith.science/pith/L7A2XEMS4GGY42BN5WH7AMKHZD.json","view_paper":"https://pith.science/paper/L7A2XEMS","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2107.02561&json=true","fetch_graph":"https://pith.science/api/pith-number/L7A2XEMS4GGY42BN5WH7AMKHZD/graph.json","fetch_events":"https://pith.science/api/pith-number/L7A2XEMS4GGY42BN5WH7AMKHZD/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/L7A2XEMS4GGY42BN5WH7AMKHZD/action/timestamp_anchor","attest_storage":"https://pith.science/pith/L7A2XEMS4GGY42BN5WH7AMKHZD/action/storage_attestation","attest_author":"https://pith.science/pith/L7A2XEMS4GGY42BN5WH7AMKHZD/action/author_attestation","sign_citation":"https://pith.science/pith/L7A2XEMS4GGY42BN5WH7AMKHZD/action/citation_signature","submit_replication":"https://pith.science/pith/L7A2XEMS4GGY42BN5WH7AMKHZD/action/replication_record"}},"created_at":"2026-07-05T03:21:54.156157+00:00","updated_at":"2026-07-05T03:21:54.156157+00:00"}