AI Paper Atlas

Research map / Large language models / Retrieval and knowledge

MoshiRAG: Asynchronous Knowledge Retrieval for Full-Duplex Speech Language Models

Chung-Ming Chien, Manu Orsini, Eugene Kharitonov, Neil Zeghidour, Karen Livescu, Alexandre Défossez

ICML 2026 · Large language models › Retrieval and knowledge · approach: Retrieval-augmented reasoning · 0 citations (OpenAlex)

Read the paperCheck your idea against 137,217 papers

AI-extracted. The measured effects below were read by AI from the paper's full text (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.

Measured effects reported in the paper

What was changedWhereMeasuredEffect
add asynchronous retrieval (MoshiRAG with Gemma 3 27B back end) to MoshiMoshi 7B · HaluEval audioaccuracy (%): 10.5 -> 36.3Improved
fine-tune Moshi on RAG data without retrieval (ablation)Vanilla Moshi · TriviaQAaccuracy (%): 22.8 -> 29.7Improved
swap retrieval back end from Gemma 3 27B to GPT-4.1 (ablation)MoshiRAG · HaluEval audioaccuracy (%): 36.3 -> 51.3Improved
swap retrieval back end from Gemma 3 27B to Tavily web search (ablation)MoshiRAG · TriviaQAaccuracy (%): 69.6 -> 77.5Improved
add RAG conversational template (lead portion before keyword)MoshiRAG vs vanilla Moshi · keyword delayE2EKD (s): 2.1 -> 3.1Hurt
fine-tune on RAG data and add retrievalMoshi · Full-Duplex-Bench PauseTOR: 0.99 -> 0.32Improved
fine-tune on RAG data with adversarial and distracting subsets (v2, v3)Moshi · Full-Duplex-Bench User InterruptionGPT score: 0.77 -> 3.75Improved
summarize LLM reference before injecting it (ablation)MoshiRAG · GSM8K (spoken)accuracy (%): 33.9 -> 51.2Improved
use MoshiRAG with GPT-4.1 back end for out-of-domain math (ablation)MoshiRAG · GSM8K (spoken)accuracy (%): 33.9 -> 43.2Improved
retrieved reference vs final spoken responseMoshiRAG · QA benchmarksaccuracy (%): 83.0 -> 80.3 (LlamaQ)Hurt

Closest papers by meaning

Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.

Open scime draft

Get the next scime Dispatch and a monthly field update

Free. We send a confirmation link first; every email has a one-click unsubscribe. Privacy