{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2025:IP2BH3AHR6BC5BF5JJJTUTGCWU","short_pith_number":"pith:IP2BH3AH","schema_version":"1.0","canonical_sha256":"43f413ec078f822e84bd4a533a4cc2b51a6fcb8da4121a0dc0d8b4fb9c48f80b","source":{"kind":"arxiv","id":"2505.02625","version":1},"attestation_state":"computed","paper":{"title":"LLaMA-Omni2: LLM-based Real-time Spoken Chatbot with Autoregressive Streaming Speech Synthesis","license":"http://creativecommons.org/licenses/by-nc-nd/4.0/","headline":"","cross_cats":["cs.AI","cs.SD","eess.AS"],"primary_cat":"cs.CL","authors_text":"Qingkai Fang, Shaolei Zhang, Shoutao Guo, Yang Feng, Yan Zhou","submitted_at":"2025-05-05T12:53:09Z","abstract_excerpt":"Real-time, intelligent, and natural speech interaction is an essential part of the next-generation human-computer interaction. Recent advancements have showcased the potential of building intelligent spoken chatbots based on large language models (LLMs). In this paper, we introduce LLaMA-Omni 2, a series of speech language models (SpeechLMs) ranging from 0.5B to 14B parameters, capable of achieving high-quality real-time speech interaction. LLaMA-Omni 2 is built upon the Qwen2.5 series models, integrating a speech encoder and an autoregressive streaming speech decoder. Despite being trained on"},"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":"2505.02625","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/licenses/by-nc-nd/4.0/","primary_cat":"cs.CL","submitted_at":"2025-05-05T12:53:09Z","cross_cats_sorted":["cs.AI","cs.SD","eess.AS"],"title_canon_sha256":"8dd3d402af0da5d5e8ed235fb95c2563d7d96e345f4e184f489a53ad101dc8e6","abstract_canon_sha256":"c9cbb641fe5c125435bca5e5a296757ad21e6a6b36fb197def8d46a6fd593935"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T10:58:40.519444Z","signature_b64":"xNVwQfkWwYW3t50LTl+JDNNwpCb1qZNxYGJMqpra9n854L8rRTwpA/4/ZOEdk0FcgNfzoiCS0r13QQT65Q4CAA==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"43f413ec078f822e84bd4a533a4cc2b51a6fcb8da4121a0dc0d8b4fb9c48f80b","last_reissued_at":"2026-07-05T10:58:40.519042Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T10:58:40.519042Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"LLaMA-Omni2: LLM-based Real-time Spoken Chatbot with Autoregressive Streaming Speech Synthesis","license":"http://creativecommons.org/licenses/by-nc-nd/4.0/","headline":"","cross_cats":["cs.AI","cs.SD","eess.AS"],"primary_cat":"cs.CL","authors_text":"Qingkai Fang, Shaolei Zhang, Shoutao Guo, Yang Feng, Yan Zhou","submitted_at":"2025-05-05T12:53:09Z","abstract_excerpt":"Real-time, intelligent, and natural speech interaction is an essential part of the next-generation human-computer interaction. Recent advancements have showcased the potential of building intelligent spoken chatbots based on large language models (LLMs). In this paper, we introduce LLaMA-Omni 2, a series of speech language models (SpeechLMs) ranging from 0.5B to 14B parameters, capable of achieving high-quality real-time speech interaction. LLaMA-Omni 2 is built upon the Qwen2.5 series models, integrating a speech encoder and an autoregressive streaming speech decoder. Despite being trained on"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2505.02625","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/2505.02625/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":"2505.02625","created_at":"2026-07-05T10:58:40.519099+00:00"},{"alias_kind":"arxiv_version","alias_value":"2505.02625v1","created_at":"2026-07-05T10:58:40.519099+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2505.02625","created_at":"2026-07-05T10:58:40.519099+00:00"},{"alias_kind":"pith_short_12","alias_value":"IP2BH3AHR6BC","created_at":"2026-07-05T10:58:40.519099+00:00"},{"alias_kind":"pith_short_16","alias_value":"IP2BH3AHR6BC5BF5","created_at":"2026-07-05T10:58:40.519099+00:00"},{"alias_kind":"pith_short_8","alias_value":"IP2BH3AH","created_at":"2026-07-05T10:58:40.519099+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":10,"internal_anchor_count":0,"sample":[{"citing_arxiv_id":"2606.25444","citing_title":"Does Translation-Enhanced Speech Encoder Pre-training Affect Speech LLMs?","ref_index":26,"is_internal_anchor":false},{"citing_arxiv_id":"2606.07433","citing_title":"Watch, Remember, Reason: Human-View Video Understanding with MLLMs","ref_index":127,"is_internal_anchor":false},{"citing_arxiv_id":"2606.05121","citing_title":"Audio Interaction Model","ref_index":12,"is_internal_anchor":false},{"citing_arxiv_id":"2606.30944","citing_title":"Preserving Speech-to-Text LLM Capabilities in Speech-to-Speech Generation","ref_index":29,"is_internal_anchor":false},{"citing_arxiv_id":"2606.01016","citing_title":"PolySpeech-100: A Large-Scale Benchmark for Speech Understanding Across 100+ Languages and Dialects","ref_index":35,"is_internal_anchor":false},{"citing_arxiv_id":"2509.22220","citing_title":"StableToken: A Noise-Robust Semantic Speech Tokenizer for Resilient SpeechLLMs","ref_index":27,"is_internal_anchor":false},{"citing_arxiv_id":"2510.09592","citing_title":"Mind-Paced Speaking: A Dual-Brain Approach to Real-Time Reasoning in Spoken Language Models","ref_index":18,"is_internal_anchor":false},{"citing_arxiv_id":"2604.19300","citing_title":"HalluAudio: A Comprehensive Benchmark for Hallucination Detection in Large Audio-Language Models","ref_index":8,"is_internal_anchor":false},{"citing_arxiv_id":"2604.13804","citing_title":"Character Beyond Speech: Leveraging Role-Playing Evaluation in Audio Large Language Models via Reinforcement Learning","ref_index":12,"is_internal_anchor":false},{"citing_arxiv_id":"2604.20842","citing_title":"SpeechParaling-Bench: A Comprehensive Benchmark for Paralinguistic-Aware Speech Generation","ref_index":11,"is_internal_anchor":false}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/IP2BH3AHR6BC5BF5JJJTUTGCWU","json":"https://pith.science/pith/IP2BH3AHR6BC5BF5JJJTUTGCWU.json","graph_json":"https://pith.science/api/pith-number/IP2BH3AHR6BC5BF5JJJTUTGCWU/graph.json","events_json":"https://pith.science/api/pith-number/IP2BH3AHR6BC5BF5JJJTUTGCWU/events.json","paper":"https://pith.science/paper/IP2BH3AH"},"agent_actions":{"view_html":"https://pith.science/pith/IP2BH3AHR6BC5BF5JJJTUTGCWU","download_json":"https://pith.science/pith/IP2BH3AHR6BC5BF5JJJTUTGCWU.json","view_paper":"https://pith.science/paper/IP2BH3AH","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2505.02625&json=true","fetch_graph":"https://pith.science/api/pith-number/IP2BH3AHR6BC5BF5JJJTUTGCWU/graph.json","fetch_events":"https://pith.science/api/pith-number/IP2BH3AHR6BC5BF5JJJTUTGCWU/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/IP2BH3AHR6BC5BF5JJJTUTGCWU/action/timestamp_anchor","attest_storage":"https://pith.science/pith/IP2BH3AHR6BC5BF5JJJTUTGCWU/action/storage_attestation","attest_author":"https://pith.science/pith/IP2BH3AHR6BC5BF5JJJTUTGCWU/action/author_attestation","sign_citation":"https://pith.science/pith/IP2BH3AHR6BC5BF5JJJTUTGCWU/action/citation_signature","submit_replication":"https://pith.science/pith/IP2BH3AHR6BC5BF5JJJTUTGCWU/action/replication_record"}},"created_at":"2026-07-05T10:58:40.519099+00:00","updated_at":"2026-07-05T10:58:40.519099+00:00"}