跳到正文
HuggingFace Daily Papers(社區熱門論文)·· 2026-10-02AI 評分51

DuplexJev 不經解碼即可批次處理語音決策,單 Token 監督讓凍結 LLM 聽見逐字稿以外的資訊

Batched Speech Decisions Without Decoding: Single-Token Supervision Lets a Frozen LLM Hear Beyond the Transcript

AI 導讀

DuplexJev 提出將 ASR 編碼器的隱藏狀態經小型連接器送入凍結的 LLM,直接以單個 Token 對選項作分佈預測,無需自回歸解碼。在 8-GPU 節點上,系統約 0.1 s 內可針對八段語音作出 80 項決策;使用最後一層連接器時,語音問答準確率為 90%,接近讀取逐字稿的 91%。

正文

Published on Oct 2

Authors:

,

,

,

,

,

Abstract

Full-duplex voice agents make many small, closed decisions, which current systems answer by slow autoregressive decoding. We propose DuplexJev, which feeds ASR-encoder hidden states through a small connector into a frozen LLM and reads each question as a single-token distribution over its options. Nothing is decoded, and an 8-GPU node answers 80 decisions about eight utterances in about 0.1 s. With a last-layer connector, spoken QA stays close to reading the transcript (90% vs. 91%). DuplexJev also hears the speaker: gender and emotion accuracy both reach 90% (from 55% and 28%) with a cross-attention connector, whose spoken QA drops by only 1 point (83% to 82%). We train decisions with cross-entropy on the read-out answer token, instead of the usual transcript distillation, whose teacher never hears the voice, and keep distillation for content. Encoders and LLMs are interchangeable; we release weights, training recipe, a batched-inference pipeline for full-duplex serving and a bilingual spoken-QA set.

View arXiv page View PDF Add to collection

Get this paper in your agent:

hf papers read 2610.02638

Don't have the latest CLI?

curl -LsSf https://hf.co/cli/install.sh | bash

Models citing this paper 11

adventists-ai/DuplexJev-4B

Audio-Text-to-Text •

4B •

Updated about 23 hours ago

•

80

•

1

adventists-ai/DuplexJev-32B

Audio-Text-to-Text •

33B •

Updated about 23 hours ago

•

38

adventists-ai/DuplexJev-Gemma-31B

Audio-Text-to-Text •

31B •

Updated about 23 hours ago

•

27

adventists-ai/DuplexJev-4B-Turn

Audio-Text-to-Text •

4B •

Updated about 23 hours ago

•

40

Browse 11 models citing this paper

Datasets citing this paper 0

No dataset linking this paper

Cite arxiv.org/abs/2610.02638 in a dataset README.md to link it from this page.

Spaces citing this paper 3

Collections including this paper 1

來源:HuggingFace Daily Papers(社區熱門論文) · huggingface.co