# Learning Natural Conversational Behavior in Tandem Speech-to-Speech Models with Randomized Guidance

arXiv: https://arxiv.org/abs/2609.30773 · HTML: https://mlsanity.com/paper/2609.30773/

## Key facts

- No citing papers in the map yet (data through 14 Oct 2026).
- Builds on 14 of its 15 references that are in the map.
- First posted on arXiv 25 Sep 2026.

## Builds on (most cited first, 14 of 14)

- [Judging LLM-as-a-Judge with MT-Bench and Chatbot Arena](https://mlsanity.com/paper/2306.05685/) (2023)
- [Robust Speech Recognition via Large-Scale Weak Supervision](https://mlsanity.com/paper/2212.04356/) (2022)
- [MiniLM: Deep Self-Attention Distillation for Task-Agnostic Compression of Pre-Trained Transformers](https://mlsanity.com/paper/2002.10957/) (2020)
- [Moshi: a speech-text foundation model for real-time dialogue](https://mlsanity.com/paper/2410.00037/) (2024)
- [DNSMOS: A Non-Intrusive Perceptual Objective Speech Quality metric to evaluate Noise Suppressors](https://mlsanity.com/paper/2010.15258/) (2020)
- [MTR-DuplexBench: Towards a Comprehensive Evaluation of Multi-Round Conversations for Full-Duplex Speech Language Models](https://mlsanity.com/paper/2511.10262/) (2025)
- [Audio-Aware Large Language Models as Judges for Speaking Styles](https://mlsanity.com/paper/2506.05984/) (2025)
- [MoshiRAG: Asynchronous Knowledge Retrieval for Full-Duplex Speech Language Models](https://mlsanity.com/paper/2604.12928/) (2026)
- [Multi-Faceted Interactivity Alignment in Full-Duplex Speech Models](https://mlsanity.com/paper/2606.11167/) (2026)
- [Sidon: Fast and Robust Open-Source Multilingual Speech Restoration for Large-scale Dataset Cleansing](https://mlsanity.com/paper/2509.17052/) (2025)
- [KAME: Tandem Architecture for Enhancing Knowledge in Real-Time Speech-to-Speech Conversational AI](https://mlsanity.com/paper/2510.02327/) (2025)
- [DuplexChat: Constructing Speaker-Separated Full-Duplex Dialogue Speech at Scale for Spoken Dialogue Language Modeling](https://mlsanity.com/paper/2607.04941/) (2026)
- [Thinking While Speaking: Inference-Time Knowledge Transfer for Responsive and Intelligent Conversational Voice Agents](https://mlsanity.com/paper/2511.07397/) (2025)
- [ConversationalVoice: Full-Duplex Speech Data from Real Conversations through Source-Faithful Reconstruction and Conversation-Grounded Expansion](https://mlsanity.com/paper/2609.08147/) (2026)

## Abstract (arXiv)

Tandem speech-to-speech architectures couple a responsive speech frontend with an asynchronous text backend. In KAME, a large language model (LLM) serves as the backend, supplying candidate responses as guidance to the speech frontend while the user is still speaking. Ordinary conversation recordings capture the eventual response but not the guidance the backend would supply during the user's utterance. Generating the missing guidance with a simulator LLM adds substantial data-preparation overhead when training on real conversations. We propose randomized intermediate guidance, which derives guidance directly from the conversation corpus rather than simulating backend LLM behavior. During training, target responses provide informative guidance, while randomly sampled responses provide potentially irrelevant updates during the utterance. This combination aims to teach the frontend to use backend information selectively. On synthetic dialogues, KAME trained with this recipe achieves response quality comparable to that of the LLM-generated and similarity-based baselines. Training on 3.8k hours of real conversations improves smooth turn-taking and audio-judge naturalness over synthetic-data KAME while retaining a response-quality advantage over Moshi. These results show that randomized guidance offers a practical route to combining the response-quality benefits of tandem models with natural conversational behavior learned from real speech.