作者丨 幸丽娟 编辑丨岑 峰  
（翻译）使用 Sentence Transformers 的多向量（后期交互）嵌入模型
We’re on a journey to advance and democratize artificial intelligence through open source and open science.
（翻译）构建低延迟多语言语音智能体：借助 NVIDIA Magpie TTS 实现开放权重与完整部署控制
A Blog post by NVIDIA on Hugging Face
（翻译）Gemini 3.5 Live Translate：流畅自然的语音翻译
Gemini 3.5 Live Translate brings near real-time, natural speech translation to Google AI Studio, Google Translate and Google Meet.
（翻译）使用 Granite Speech 5.0 Turbo CTC 实现极其快速且准确的转录
A Blog post by IBM Granite on Hugging Face
（翻译）MiNER：面向临床文本中疟疾疾病实体识别的微调生物医学自然语言处理
Abstract page for arXiv paper 2609.00073: MiNER: Fine-Tuned Biomedical Natural Language Processing for Malaria Disease Entity Recognition in Clinical Texts
（翻译）陷入故事：多轮LLM对话中的叙事俘获
Abstract page for arXiv paper 2609.03407: Caught in the Story: Narrative Captivity in Multi-turn LLMs Conversation
（翻译）DuplexSpeechBench-IFEval：评估全双工语音智能体中的隐式指令遵循
Abstract page for arXiv paper 2609.03423: DuplexSpeechBench-IFEval: Evaluating Implicit Instruction Following in Full-Duplex Voice Agents
Google为多个软件新增Gemini语音功能；华为新款三折叠手机搭载麒麟9050 Pro芯片和HarmonyOS 7系统；科大讯飞发布星火Spark-X2.5大模型
（翻译）RESCUE-BENCH：面向关系感知的多方情感支持对话系统
Abstract page for arXiv paper 2609.09657: RESCUE-BENCH: Towards Relation-Aware Multi-Party Emotional Support Conversation Systems
（翻译）MOSAIC：面向查询感知的 GraphRAG 探索策略适配
arXiv:2609.11065v1 Announce Type: new Abstract: Graph Retrieval-Augmented Generation (GraphRAG) can connect evidence distributed across a corpus graph, but most systems use largely shared exploration procedures across queries. This creates a structural mismatch: direct facts may need compact local n
（翻译）KuaiRP 系列角色扮演模型技术报告
Abstract page for arXiv paper 2609.11127: KuaiRP Series Role-playing Models Technical Report
9月15日，阶跃发布新一代语音大模型 StepAudio 3 系列，一次性推出 StepAudio 3 Realtime、StepAudio 3 ASR、StepAudio 3 TTS、StepAudio 3 Gen 和 StepAudio 3 Music 五款模型，覆盖实时语音交互、语音识别、真人级语音生成和音乐创作等场景。其中，多款模型在 权威 第三方 AI 模型评测机构 Artificial Analysis 榜单中位列全球第一。
（翻译）让 AI 覆盖每一种语言和每一个人
We’re moving beyond traditional text translation to build models that understand the world’s rich, living languages exactly as they are expressed.
