{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2026:TLSSVJTVQQDEK7RLNDTGSPIZCR","short_pith_number":"pith:TLSSVJTV","schema_version":"1.0","canonical_sha256":"9ae52aa6758406457e2b68e6693d191446f3137a1fbdf16e6ac3a7762396b3d9","source":{"kind":"arxiv","id":"2601.03790","version":4},"attestation_state":"computed","paper":{"title":"NeoAMT: Neologism-Aware Agentic Machine Translation with Reinforcement Learning","license":"http://creativecommons.org/licenses/by-nc-sa/4.0/","headline":"An RL agent with a Wiktionary search toolkit translates neologisms more accurately across 75 directions.","cross_cats":["cs.AI"],"primary_cat":"cs.CL","authors_text":"Kaiyan Zhao, Masaaki Nagata, Yoshimasa Tsuruoka, Zhongtao Miao","submitted_at":"2026-01-07T10:49:00Z","abstract_excerpt":"Neologism-aware machine translation aims to translate source sentences containing neologisms into target languages. This field remains underexplored compared with general machine translation (MT). In this paper, we propose an agentic framework, NeoAMT, for neologism-aware machine translation equipped with a Wiktionary-based search toolkit. Specifically, we first construct a dedicated dataset for neologism-aware machine translation and build a search toolkit grounded in Wiktionary. The dataset covers 16 languages and 75 translation directions in total, derived from approximately 10 million reco"},"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":true},"canonical_record":{"source":{"id":"2601.03790","kind":"arxiv","version":4},"metadata":{"license":"http://creativecommons.org/licenses/by-nc-sa/4.0/","primary_cat":"cs.CL","submitted_at":"2026-01-07T10:49:00Z","cross_cats_sorted":["cs.AI"],"title_canon_sha256":"7b68cce7251b6703597100162d5ef4a3284d1a39c70afca2f6f1dd688d86c439","abstract_canon_sha256":"4881e13c34a555e115313a4a4158c51c4b9a8a3b3ae0a27ec366fe8cd2294ba3"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-05-26T02:04:02.427392Z","signature_b64":"O9FwDliM7oUcpdJArdwBXcR+dbxO2VbKb2iuIuefANml1WmDuCx0LpN///F7wQlzd0XVhBVMpu1VsC1NpkTTCA==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"9ae52aa6758406457e2b68e6693d191446f3137a1fbdf16e6ac3a7762396b3d9","last_reissued_at":"2026-05-26T02:04:02.426585Z","signature_status":"signed_v1","first_computed_at":"2026-05-26T02:04:02.426585Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"NeoAMT: Neologism-Aware Agentic Machine Translation with Reinforcement Learning","license":"http://creativecommons.org/licenses/by-nc-sa/4.0/","headline":"An RL agent with a Wiktionary search toolkit translates neologisms more accurately across 75 directions.","cross_cats":["cs.AI"],"primary_cat":"cs.CL","authors_text":"Kaiyan Zhao, Masaaki Nagata, Yoshimasa Tsuruoka, Zhongtao Miao","submitted_at":"2026-01-07T10:49:00Z","abstract_excerpt":"Neologism-aware machine translation aims to translate source sentences containing neologisms into target languages. This field remains underexplored compared with general machine translation (MT). In this paper, we propose an agentic framework, NeoAMT, for neologism-aware machine translation equipped with a Wiktionary-based search toolkit. Specifically, we first construct a dedicated dataset for neologism-aware machine translation and build a search toolkit grounded in Wiktionary. The dataset covers 16 languages and 75 translation directions in total, derived from approximately 10 million reco"},"claims":{"count":4,"items":[{"kind":"strongest_claim","text":"We propose an RL training framework featuring a novel reward design and an adaptive rollout generation strategy that exploits translation difficulty to further improve the translation quality of translation agents using our search toolkit.","source":"verdict.strongest_claim","status":"machine_extracted","claim_id":"C1","attestation":"unclaimed"},{"kind":"weakest_assumption","text":"That the Wiktionary-derived dataset and search toolkit supply sufficiently accurate and comprehensive information for the RL agent to learn reliable neologism-handling policies that generalize across the 75 translation directions.","source":"verdict.weakest_assumption","status":"machine_extracted","claim_id":"C2","attestation":"unclaimed"},{"kind":"one_line_summary","text":"NeoAMT trains an RL agent with a Wiktionary toolkit, novel reward, and adaptive rollouts to translate sentences containing neologisms in 16 languages and 75 directions.","source":"verdict.one_line_summary","status":"machine_extracted","claim_id":"C3","attestation":"unclaimed"},{"kind":"headline","text":"An RL agent with a Wiktionary search toolkit translates neologisms more accurately across 75 directions.","source":"verdict.pith_extraction.headline","status":"machine_extracted","claim_id":"C4","attestation":"unclaimed"}],"snapshot_sha256":"ba7dff8bb59ff8c57dd53b71d74b38ca0b517b7fbe9b0d7978579a841fe1fd0d"},"source":{"id":"2601.03790","kind":"arxiv","version":4},"verdict":{"id":"4cb2b2b2-be80-451d-b341-a0fbdf7adc96","model_set":{"reader":"grok-4.3"},"created_at":"2026-05-16T16:38:57.984909Z","strongest_claim":"We propose an RL training framework featuring a novel reward design and an adaptive rollout generation strategy that exploits translation difficulty to further improve the translation quality of translation agents using our search toolkit.","one_line_summary":"NeoAMT trains an RL agent with a Wiktionary toolkit, novel reward, and adaptive rollouts to translate sentences containing neologisms in 16 languages and 75 directions.","pipeline_version":"pith-pipeline@v0.9.0","weakest_assumption":"That the Wiktionary-derived dataset and search toolkit supply sufficiently accurate and comprehensive information for the RL agent to learn reliable neologism-handling policies that generalize across the 75 translation directions.","pith_extraction_headline":"An RL agent with a Wiktionary search toolkit translates neologisms more accurately across 75 directions."},"integrity":{"clean":true,"summary":{"advisory":0,"critical":0,"by_detector":{},"informational":0},"endpoint":"/pith/2601.03790/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":2,"snapshot_sha256":"f73f0435853cae26a7eb921ec6ab4a56d77a9848f7d2e0fdb830ea18ba8fee7b"},"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":"2601.03790","created_at":"2026-05-26T02:04:02.426731+00:00"},{"alias_kind":"arxiv_version","alias_value":"2601.03790v4","created_at":"2026-05-26T02:04:02.426731+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2601.03790","created_at":"2026-05-26T02:04:02.426731+00:00"},{"alias_kind":"pith_short_12","alias_value":"TLSSVJTVQQDE","created_at":"2026-05-26T02:04:02.426731+00:00"},{"alias_kind":"pith_short_16","alias_value":"TLSSVJTVQQDEK7RL","created_at":"2026-05-26T02:04:02.426731+00:00"},{"alias_kind":"pith_short_8","alias_value":"TLSSVJTV","created_at":"2026-05-26T02:04:02.426731+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":2,"internal_anchor_count":2,"sample":[{"citing_arxiv_id":"2605.09100","citing_title":"GRC: Unifying Reasoning-Driven Generation, Retrieval and Compression","ref_index":27,"is_internal_anchor":true},{"citing_arxiv_id":"2605.09100","citing_title":"GRC: Unifying Reasoning-Driven Generation, Retrieval and Compression","ref_index":27,"is_internal_anchor":true}]},"formal_canon":{"evidence_count":2,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/TLSSVJTVQQDEK7RLNDTGSPIZCR","json":"https://pith.science/pith/TLSSVJTVQQDEK7RLNDTGSPIZCR.json","graph_json":"https://pith.science/api/pith-number/TLSSVJTVQQDEK7RLNDTGSPIZCR/graph.json","events_json":"https://pith.science/api/pith-number/TLSSVJTVQQDEK7RLNDTGSPIZCR/events.json","paper":"https://pith.science/paper/TLSSVJTV"},"agent_actions":{"view_html":"https://pith.science/pith/TLSSVJTVQQDEK7RLNDTGSPIZCR","download_json":"https://pith.science/pith/TLSSVJTVQQDEK7RLNDTGSPIZCR.json","view_paper":"https://pith.science/paper/TLSSVJTV","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2601.03790&json=true","fetch_graph":"https://pith.science/api/pith-number/TLSSVJTVQQDEK7RLNDTGSPIZCR/graph.json","fetch_events":"https://pith.science/api/pith-number/TLSSVJTVQQDEK7RLNDTGSPIZCR/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/TLSSVJTVQQDEK7RLNDTGSPIZCR/action/timestamp_anchor","attest_storage":"https://pith.science/pith/TLSSVJTVQQDEK7RLNDTGSPIZCR/action/storage_attestation","attest_author":"https://pith.science/pith/TLSSVJTVQQDEK7RLNDTGSPIZCR/action/author_attestation","sign_citation":"https://pith.science/pith/TLSSVJTVQQDEK7RLNDTGSPIZCR/action/citation_signature","submit_replication":"https://pith.science/pith/TLSSVJTVQQDEK7RLNDTGSPIZCR/action/replication_record"}},"created_at":"2026-05-26T02:04:02.426731+00:00","updated_at":"2026-05-26T02:04:02.426731+00:00"}