{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2025:ZK42YHBLJOQURCX46TAGHVQIEQ","short_pith_number":"pith:ZK42YHBL","schema_version":"1.0","canonical_sha256":"cab9ac1c2b4ba1488afcf4c063d6082439f54c95e24872b1939d2e69dfa8bfa0","source":{"kind":"arxiv","id":"2507.00425","version":1},"attestation_state":"computed","paper":{"title":"Flexible Language Modeling in Continuous Space with Transformer-based Autoregressive Flows","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.CL"],"primary_cat":"cs.LG","authors_text":"Huangjie Zheng, Jiatao Gu, Josh Susskind, Miguel Angel Bautista, Navdeep Jaitly, Ruixiang Zhang, Shuangfei Zhai, Tianrong Chen, Yizhe Zhang","submitted_at":"2025-07-01T04:51:25Z","abstract_excerpt":"Autoregressive models have driven remarkable progress in language modeling. Their foundational reliance on discrete tokens, unidirectional context, and single-pass decoding, while central to their success, also inspires the exploration of a design space that could offer new axes of modeling flexibility. In this work, we explore an alternative paradigm, shifting language modeling from a discrete token space to a continuous latent space. We propose a novel framework TarFlowLM, that employs transformer-based autoregressive normalizing flows to model these continuous representations. This approach"},"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":"2507.00425","kind":"arxiv","version":1},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.LG","submitted_at":"2025-07-01T04:51:25Z","cross_cats_sorted":["cs.CL"],"title_canon_sha256":"ed129daf79df56def0bd8ce18f84b3dfac063477004f5b6818dd19cb9f3a175b","abstract_canon_sha256":"a6e95ab0824feb1788c8556a47c4d2c418e6648b63e4bd9e8cde3627bd6d3cf9"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T11:30:08.680185Z","signature_b64":"v1AUynhHUSq2mF33bZgIHGqPgbEWp7teAMjg6AC1dExL+CWx06uJh/POlbiwRxCA0xNBzqDAeozjtGxVO1hEDQ==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"cab9ac1c2b4ba1488afcf4c063d6082439f54c95e24872b1939d2e69dfa8bfa0","last_reissued_at":"2026-07-05T11:30:08.679730Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T11:30:08.679730Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Flexible Language Modeling in Continuous Space with Transformer-based Autoregressive Flows","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.CL"],"primary_cat":"cs.LG","authors_text":"Huangjie Zheng, Jiatao Gu, Josh Susskind, Miguel Angel Bautista, Navdeep Jaitly, Ruixiang Zhang, Shuangfei Zhai, Tianrong Chen, Yizhe Zhang","submitted_at":"2025-07-01T04:51:25Z","abstract_excerpt":"Autoregressive models have driven remarkable progress in language modeling. Their foundational reliance on discrete tokens, unidirectional context, and single-pass decoding, while central to their success, also inspires the exploration of a design space that could offer new axes of modeling flexibility. In this work, we explore an alternative paradigm, shifting language modeling from a discrete token space to a continuous latent space. We propose a novel framework TarFlowLM, that employs transformer-based autoregressive normalizing flows to model these continuous representations. This approach"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2507.00425","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/2507.00425/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":"2507.00425","created_at":"2026-07-05T11:30:08.679786+00:00"},{"alias_kind":"arxiv_version","alias_value":"2507.00425v1","created_at":"2026-07-05T11:30:08.679786+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2507.00425","created_at":"2026-07-05T11:30:08.679786+00:00"},{"alias_kind":"pith_short_12","alias_value":"ZK42YHBLJOQU","created_at":"2026-07-05T11:30:08.679786+00:00"},{"alias_kind":"pith_short_16","alias_value":"ZK42YHBLJOQURCX4","created_at":"2026-07-05T11:30:08.679786+00:00"},{"alias_kind":"pith_short_8","alias_value":"ZK42YHBL","created_at":"2026-07-05T11:30:08.679786+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":1,"internal_anchor_count":0,"sample":[{"citing_arxiv_id":"2604.20041","citing_title":"Normalizing Flows with Iterative Denoising","ref_index":23,"is_internal_anchor":false}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/ZK42YHBLJOQURCX46TAGHVQIEQ","json":"https://pith.science/pith/ZK42YHBLJOQURCX46TAGHVQIEQ.json","graph_json":"https://pith.science/api/pith-number/ZK42YHBLJOQURCX46TAGHVQIEQ/graph.json","events_json":"https://pith.science/api/pith-number/ZK42YHBLJOQURCX46TAGHVQIEQ/events.json","paper":"https://pith.science/paper/ZK42YHBL"},"agent_actions":{"view_html":"https://pith.science/pith/ZK42YHBLJOQURCX46TAGHVQIEQ","download_json":"https://pith.science/pith/ZK42YHBLJOQURCX46TAGHVQIEQ.json","view_paper":"https://pith.science/paper/ZK42YHBL","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2507.00425&json=true","fetch_graph":"https://pith.science/api/pith-number/ZK42YHBLJOQURCX46TAGHVQIEQ/graph.json","fetch_events":"https://pith.science/api/pith-number/ZK42YHBLJOQURCX46TAGHVQIEQ/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/ZK42YHBLJOQURCX46TAGHVQIEQ/action/timestamp_anchor","attest_storage":"https://pith.science/pith/ZK42YHBLJOQURCX46TAGHVQIEQ/action/storage_attestation","attest_author":"https://pith.science/pith/ZK42YHBLJOQURCX46TAGHVQIEQ/action/author_attestation","sign_citation":"https://pith.science/pith/ZK42YHBLJOQURCX46TAGHVQIEQ/action/citation_signature","submit_replication":"https://pith.science/pith/ZK42YHBLJOQURCX46TAGHVQIEQ/action/replication_record"}},"created_at":"2026-07-05T11:30:08.679786+00:00","updated_at":"2026-07-05T11:30:08.679786+00:00"}