{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2026:LATP7XZ3YN7KTY3ROLEY4EOW2G","short_pith_number":"pith:LATP7XZ3","schema_version":"1.0","canonical_sha256":"5826ffdf3bc37ea9e37172c98e11d6d1bf71d8f1285ee96488a30523283d5a67","source":{"kind":"arxiv","id":"2605.27255","version":1},"attestation_state":"computed","paper":{"title":"Pair-In, Pair-Out: Latent Multi-Token Prediction for Efficient LLMs","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.AI"],"primary_cat":"cs.CL","authors_text":"Liujie Zhang, Minghao Li, Ruihua Song, Siqi Fan, Weihang Chen, Wenhui Tan, Xiaoqian Ma, Xiusheng Huang","submitted_at":"2026-05-26T16:31:45Z","abstract_excerpt":"Long chain-of-thought reasoning has made autoregressive decoding the dominant inference cost of modern large language models. Existing methods target either the input side (latent compression) or the output side (speculative decoding and multi-token prediction, MTP), but the two lines of work have been pursued independently. Moreover, output-side methods must incur an expensive verifier pass to validate the unreliable draft tokens predicted by MTP. To address these issues, we propose \\textbf{Pair-In, Pair-Out (PIPO)}, which unifies both sides by viewing a latent compressor and an MTP head as m"},"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":"2605.27255","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CL","submitted_at":"2026-05-26T16:31:45Z","cross_cats_sorted":["cs.AI"],"title_canon_sha256":"27e1ea46da98df117037c75f7c52ac9f446f78a24528e98d305566b8506b4248","abstract_canon_sha256":"6de9fe727c6101828ed8f4eb3d21a0fdbd28d37892a352f6dc0539379d9e61e5"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-05-27T02:06:08.624240Z","signature_b64":"05+tV6BX8Nv8W2EzCnG57qptWCbbWM20WMXsgH1SlQ74x2zIDBI5uB3D/ZcLlTvauqaf1bwhTnGZsV02638kAA==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"5826ffdf3bc37ea9e37172c98e11d6d1bf71d8f1285ee96488a30523283d5a67","last_reissued_at":"2026-05-27T02:06:08.623752Z","signature_status":"signed_v1","first_computed_at":"2026-05-27T02:06:08.623752Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Pair-In, Pair-Out: Latent Multi-Token Prediction for Efficient LLMs","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.AI"],"primary_cat":"cs.CL","authors_text":"Liujie Zhang, Minghao Li, Ruihua Song, Siqi Fan, Weihang Chen, Wenhui Tan, Xiaoqian Ma, Xiusheng Huang","submitted_at":"2026-05-26T16:31:45Z","abstract_excerpt":"Long chain-of-thought reasoning has made autoregressive decoding the dominant inference cost of modern large language models. Existing methods target either the input side (latent compression) or the output side (speculative decoding and multi-token prediction, MTP), but the two lines of work have been pursued independently. Moreover, output-side methods must incur an expensive verifier pass to validate the unreliable draft tokens predicted by MTP. To address these issues, we propose \\textbf{Pair-In, Pair-Out (PIPO)}, which unifies both sides by viewing a latent compressor and an MTP head as m"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2605.27255","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/2605.27255/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":"2605.27255","created_at":"2026-05-27T02:06:08.623828+00:00"},{"alias_kind":"arxiv_version","alias_value":"2605.27255v1","created_at":"2026-05-27T02:06:08.623828+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2605.27255","created_at":"2026-05-27T02:06:08.623828+00:00"},{"alias_kind":"pith_short_12","alias_value":"LATP7XZ3YN7K","created_at":"2026-05-27T02:06:08.623828+00:00"},{"alias_kind":"pith_short_16","alias_value":"LATP7XZ3YN7KTY3R","created_at":"2026-05-27T02:06:08.623828+00:00"},{"alias_kind":"pith_short_8","alias_value":"LATP7XZ3","created_at":"2026-05-27T02:06:08.623828+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/LATP7XZ3YN7KTY3ROLEY4EOW2G","json":"https://pith.science/pith/LATP7XZ3YN7KTY3ROLEY4EOW2G.json","graph_json":"https://pith.science/api/pith-number/LATP7XZ3YN7KTY3ROLEY4EOW2G/graph.json","events_json":"https://pith.science/api/pith-number/LATP7XZ3YN7KTY3ROLEY4EOW2G/events.json","paper":"https://pith.science/paper/LATP7XZ3"},"agent_actions":{"view_html":"https://pith.science/pith/LATP7XZ3YN7KTY3ROLEY4EOW2G","download_json":"https://pith.science/pith/LATP7XZ3YN7KTY3ROLEY4EOW2G.json","view_paper":"https://pith.science/paper/LATP7XZ3","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2605.27255&json=true","fetch_graph":"https://pith.science/api/pith-number/LATP7XZ3YN7KTY3ROLEY4EOW2G/graph.json","fetch_events":"https://pith.science/api/pith-number/LATP7XZ3YN7KTY3ROLEY4EOW2G/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/LATP7XZ3YN7KTY3ROLEY4EOW2G/action/timestamp_anchor","attest_storage":"https://pith.science/pith/LATP7XZ3YN7KTY3ROLEY4EOW2G/action/storage_attestation","attest_author":"https://pith.science/pith/LATP7XZ3YN7KTY3ROLEY4EOW2G/action/author_attestation","sign_citation":"https://pith.science/pith/LATP7XZ3YN7KTY3ROLEY4EOW2G/action/citation_signature","submit_replication":"https://pith.science/pith/LATP7XZ3YN7KTY3ROLEY4EOW2G/action/replication_record"}},"created_at":"2026-05-27T02:06:08.623828+00:00","updated_at":"2026-05-27T02:06:08.623828+00:00"}