{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2022:DTAZX537WWUSSMZUKP7QNBHZCF","short_pith_number":"pith:DTAZX537","schema_version":"1.0","canonical_sha256":"1cc19bf77fb5a929333453ff0684f9114310d8bd15850c183e594d30e2babeaa","source":{"kind":"arxiv","id":"2211.09761","version":2},"attestation_state":"computed","paper":{"title":"Efficient Transformers with Dynamic Token Pooling","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":[],"primary_cat":"cs.CL","authors_text":"Adrian {\\L}a\\'ncucki, Edoardo M. Ponti, Jan Chorowski, Piotr Nawrot","submitted_at":"2022-11-17T18:39:23Z","abstract_excerpt":"Transformers achieve unrivalled performance in modelling language, but remain inefficient in terms of memory and time complexity. A possible remedy is to reduce the sequence length in the intermediate layers by pooling fixed-length segments of tokens. Nevertheless, natural units of meaning, such as words or phrases, display varying sizes. To address this mismatch, we equip language models with a dynamic-pooling mechanism, which predicts segment boundaries in an autoregressive fashion. We compare several methods to infer boundaries, including end-to-end learning through stochastic re-parameteri"},"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":"2211.09761","kind":"arxiv","version":2},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CL","submitted_at":"2022-11-17T18:39:23Z","cross_cats_sorted":[],"title_canon_sha256":"2e4d10f01d17cc813367c1d0aaf40eb9365bbf6c77ab5e1c3af251153ff7ca88","abstract_canon_sha256":"d0634c03a0d51fb926485ca9e09416d5666facbd2934e3eb02fe62dbb67a7050"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T07:04:11.648095Z","signature_b64":"vVduULX9F/GNfIDiVjwAaZYjT6POKTO1v741BCkEhVQDPd5c2o36nDon7FJP8Cs6EnvpujfGykvzvWDYf97/Bw==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"1cc19bf77fb5a929333453ff0684f9114310d8bd15850c183e594d30e2babeaa","last_reissued_at":"2026-07-05T07:04:11.647632Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T07:04:11.647632Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Efficient Transformers with Dynamic Token Pooling","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":[],"primary_cat":"cs.CL","authors_text":"Adrian {\\L}a\\'ncucki, Edoardo M. Ponti, Jan Chorowski, Piotr Nawrot","submitted_at":"2022-11-17T18:39:23Z","abstract_excerpt":"Transformers achieve unrivalled performance in modelling language, but remain inefficient in terms of memory and time complexity. A possible remedy is to reduce the sequence length in the intermediate layers by pooling fixed-length segments of tokens. Nevertheless, natural units of meaning, such as words or phrases, display varying sizes. To address this mismatch, we equip language models with a dynamic-pooling mechanism, which predicts segment boundaries in an autoregressive fashion. We compare several methods to infer boundaries, including end-to-end learning through stochastic re-parameteri"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2211.09761","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/2211.09761/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":"2211.09761","created_at":"2026-07-05T07:04:11.647696+00:00"},{"alias_kind":"arxiv_version","alias_value":"2211.09761v2","created_at":"2026-07-05T07:04:11.647696+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2211.09761","created_at":"2026-07-05T07:04:11.647696+00:00"},{"alias_kind":"pith_short_12","alias_value":"DTAZX537WWUS","created_at":"2026-07-05T07:04:11.647696+00:00"},{"alias_kind":"pith_short_16","alias_value":"DTAZX537WWUSSMZU","created_at":"2026-07-05T07:04:11.647696+00:00"},{"alias_kind":"pith_short_8","alias_value":"DTAZX537","created_at":"2026-07-05T07:04:11.647696+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":0,"internal_anchor_count":0,"sample":[]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/DTAZX537WWUSSMZUKP7QNBHZCF","json":"https://pith.science/pith/DTAZX537WWUSSMZUKP7QNBHZCF.json","graph_json":"https://pith.science/api/pith-number/DTAZX537WWUSSMZUKP7QNBHZCF/graph.json","events_json":"https://pith.science/api/pith-number/DTAZX537WWUSSMZUKP7QNBHZCF/events.json","paper":"https://pith.science/paper/DTAZX537"},"agent_actions":{"view_html":"https://pith.science/pith/DTAZX537WWUSSMZUKP7QNBHZCF","download_json":"https://pith.science/pith/DTAZX537WWUSSMZUKP7QNBHZCF.json","view_paper":"https://pith.science/paper/DTAZX537","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2211.09761&json=true","fetch_graph":"https://pith.science/api/pith-number/DTAZX537WWUSSMZUKP7QNBHZCF/graph.json","fetch_events":"https://pith.science/api/pith-number/DTAZX537WWUSSMZUKP7QNBHZCF/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/DTAZX537WWUSSMZUKP7QNBHZCF/action/timestamp_anchor","attest_storage":"https://pith.science/pith/DTAZX537WWUSSMZUKP7QNBHZCF/action/storage_attestation","attest_author":"https://pith.science/pith/DTAZX537WWUSSMZUKP7QNBHZCF/action/author_attestation","sign_citation":"https://pith.science/pith/DTAZX537WWUSSMZUKP7QNBHZCF/action/citation_signature","submit_replication":"https://pith.science/pith/DTAZX537WWUSSMZUKP7QNBHZCF/action/replication_record"}},"created_at":"2026-07-05T07:04:11.647696+00:00","updated_at":"2026-07-05T07:04:11.647696+00:00"}