ReFrame: Silence-Aware Compute Reclamation for Full-Duplex Speech Models
cs.LG
Submitted: 2026-11-01
Updated: 2026-11-09
Terminology
Sources
- Improving language models by retrieving from trillions of tokens
- Medusa: Simple LLM Inference Acceleration Framework with Multiple Decoding Heads
- Accelerating Large Language Model Decoding with Speculative Sampling
- MoshiRAG: Asynchronous Knowledge Retrieval for Full-Duplex Speech Language Models
- Moshi: a speech-text foundation model for real-time dialogue
- How Much Does Prosody Help Turn-taking? Investigations using Voice Activity Projection Models
- BayLing-Duplex: Native Full-Duplex Speech Dialogue with a Single Autoregressive LLM
- Precise Zero-Shot Dense Retrieval without Relevance Labels
- REALM: Retrieval-Augmented Language Model Pre-Training
- DuplexOmni: Real-Time Listening, Seeing, Thinking, and Speaking for Full-Duplex Interaction
- Multilingual Turn-taking Prediction Using Voice Activity Projection
- Leveraging Passage Retrieval with Generative Models for Open Domain Question Answering
- Adaptive-RAG: Learning to Adapt Retrieval-Augmented Large Language Models through Question Complexity
- Active Retrieval Augmented Generation
- Dense Passage Retrieval for Open-Domain Question Answering
- ColBERT: Efficient and Effective Passage Search via Contextualized Late Interaction over BERT
- Fast Inference from Transformers via Speculative Decoding
- HaluEval: A Large-Scale Hallucination Evaluation Benchmark for Large Language Models
- EAGLE-2: Faster Inference of Language Models with Dynamic Draft Trees
- Baichuan-Omni-1.5 Technical Report
Related papers
- A Physics-Regulated Neural Framework for Learning 3D Grain Growth Dynamics
- FlexMoE: Learning One-for-All Multi-Budget Expert Pruning for Flexible MoE Deployment
- OptiMind: Improving Small Language Models for Optimization Modeling through Formulation Guidance and Multi-Turn Learning
- Staying in the Sweet Spot: Boost Reasoning Evolution via Responsive Rollout Difficulty Manipulation
- Examining the Impact of Data and Training Methods on Emergent and Subliminal Misalignment
- Rethinking Black-Box On-Policy Distillation: A Rubric-based Analysis