{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2024:2276YT27PFF7FG6OSMO37HTV3F","short_pith_number":"pith:2276YT27","schema_version":"1.0","canonical_sha256":"d6bfec4f5f794bf29bce931dbf9e75d971c7b0ecb8772834ea3eb346fbba78de","source":{"kind":"arxiv","id":"2407.10795","version":1},"attestation_state":"computed","paper":{"title":"Multilingual Contrastive Decoding via Language-Agnostic Layers Skipping","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":[],"primary_cat":"cs.CL","authors_text":"Chris Wendler, Jiajun Chen, Shuaijie She, Shujian Huang, Sizhe Liu, Wenhao Zhu","submitted_at":"2024-07-15T15:14:01Z","abstract_excerpt":"Decoding by contrasting layers (DoLa), is designed to improve the generation quality of large language models (LLMs) by contrasting the prediction probabilities between an early exit output (amateur logits) and the final output (expert logits). However, we find that this approach does not work well on non-English tasks. Inspired by previous interpretability work on language transition during the model's forward pass, we discover that this issue arises from a language mismatch between early exit output and final output. In this work, we propose an improved contrastive decoding algorithm that is"},"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":"2407.10795","kind":"arxiv","version":1},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CL","submitted_at":"2024-07-15T15:14:01Z","cross_cats_sorted":[],"title_canon_sha256":"d404c0e6c1f2abd270fd41154e55a322edea5e708650ee990c4d576c2d1ea832","abstract_canon_sha256":"4319c8de01a27e340fc1c81ae41090e5d9df3e14f81ca9bd46b63167189d2a8b"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T08:44:04.196183Z","signature_b64":"Eotw9Cv4DGjhLH8eUMvPRNSUMS6/KtKtbUry/dgExjdyDwrHd8M1Rh+uvSIsvQpMdrz1fmO0Aka+YvAEdTwVDw==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"d6bfec4f5f794bf29bce931dbf9e75d971c7b0ecb8772834ea3eb346fbba78de","last_reissued_at":"2026-07-05T08:44:04.195632Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T08:44:04.195632Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Multilingual Contrastive Decoding via Language-Agnostic Layers Skipping","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":[],"primary_cat":"cs.CL","authors_text":"Chris Wendler, Jiajun Chen, Shuaijie She, Shujian Huang, Sizhe Liu, Wenhao Zhu","submitted_at":"2024-07-15T15:14:01Z","abstract_excerpt":"Decoding by contrasting layers (DoLa), is designed to improve the generation quality of large language models (LLMs) by contrasting the prediction probabilities between an early exit output (amateur logits) and the final output (expert logits). However, we find that this approach does not work well on non-English tasks. Inspired by previous interpretability work on language transition during the model's forward pass, we discover that this issue arises from a language mismatch between early exit output and final output. In this work, we propose an improved contrastive decoding algorithm that is"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2407.10795","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/2407.10795/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":"2407.10795","created_at":"2026-07-05T08:44:04.195692+00:00"},{"alias_kind":"arxiv_version","alias_value":"2407.10795v1","created_at":"2026-07-05T08:44:04.195692+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2407.10795","created_at":"2026-07-05T08:44:04.195692+00:00"},{"alias_kind":"pith_short_12","alias_value":"2276YT27PFF7","created_at":"2026-07-05T08:44:04.195692+00:00"},{"alias_kind":"pith_short_16","alias_value":"2276YT27PFF7FG6O","created_at":"2026-07-05T08:44:04.195692+00:00"},{"alias_kind":"pith_short_8","alias_value":"2276YT27","created_at":"2026-07-05T08:44:04.195692+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/2276YT27PFF7FG6OSMO37HTV3F","json":"https://pith.science/pith/2276YT27PFF7FG6OSMO37HTV3F.json","graph_json":"https://pith.science/api/pith-number/2276YT27PFF7FG6OSMO37HTV3F/graph.json","events_json":"https://pith.science/api/pith-number/2276YT27PFF7FG6OSMO37HTV3F/events.json","paper":"https://pith.science/paper/2276YT27"},"agent_actions":{"view_html":"https://pith.science/pith/2276YT27PFF7FG6OSMO37HTV3F","download_json":"https://pith.science/pith/2276YT27PFF7FG6OSMO37HTV3F.json","view_paper":"https://pith.science/paper/2276YT27","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2407.10795&json=true","fetch_graph":"https://pith.science/api/pith-number/2276YT27PFF7FG6OSMO37HTV3F/graph.json","fetch_events":"https://pith.science/api/pith-number/2276YT27PFF7FG6OSMO37HTV3F/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/2276YT27PFF7FG6OSMO37HTV3F/action/timestamp_anchor","attest_storage":"https://pith.science/pith/2276YT27PFF7FG6OSMO37HTV3F/action/storage_attestation","attest_author":"https://pith.science/pith/2276YT27PFF7FG6OSMO37HTV3F/action/author_attestation","sign_citation":"https://pith.science/pith/2276YT27PFF7FG6OSMO37HTV3F/action/citation_signature","submit_replication":"https://pith.science/pith/2276YT27PFF7FG6OSMO37HTV3F/action/replication_record"}},"created_at":"2026-07-05T08:44:04.195692+00:00","updated_at":"2026-07-05T08:44:04.195692+00:00"}