1
网易有道首席科学家段亦涛：让AI技术扎根业务场景
网易有道"子曰"教育大模型4.0高考数学145分，跨语种无口音语音克隆，全模态融合落地业务场景。
转载来源：网易有道 本文为量子位获授权转载，观点仅为原作者所有。
2
A Finetuned SpeechLLM for Joint Multi-Granular L2 Assessment and Natural-Language Rationales
微调语音大模型，实现多粒度第二语言评估与自然语言理由生成，教育AI新突破。
arXiv:2606.09470v1 Announce Type: new Abstract: Automated L2 speech assessment can assign proficiency labels, but often lacks interpretability. We pro…
3
FOXGLOVE: Understanding Goal-Oriented and Anchored Writing Feedback from Experts and LLMs on Argumentative Essays
研究如何利用LLM提供目标导向和锚定的写作反馈，对比专家反馈，为AI辅助议论文教学提供新视角。
arXiv:2606.06271v1 Announce Type: new Abstract: While large language models (LLMs) are increasingly used to generate writing feedback, there remains n…
4
网易有道“子曰 4”多模态模型、语音合成模型全量开源
网易有道重磅开源“子曰4”多模态模型与TTS模型，教育场景数理能力达行业顶尖，支持14种语言，单卡即可运行。
IT之家 5 月 22 日消息，网易有道今日宣布，决定将“子曰”大模型 4.0 的核心双引擎 ——“多模态模型”与“语音合成（TTS）模型”， 正式面向全球全量开源 。开发者可以免费下载、部署，并基于此进行二次开发。 此次开源的“子曰 4”多模态模型（27B 参数规模）面向教育场景，支持视觉输入的数…
5
Difficulty-Controllable Cloze Question Distractor Generation
通过控制干扰项难度提升完形填空题目质量，ACL 2026 录用论文提出新方法。
arXiv:2511.01526v2 Announce Type: replace Abstract: Multiple-choice cloze questions are commonly used to assess linguistic proficiency and comprehensi…
6
EduVQA: Towards Concept-Aware Assessment of Educational AI-Generated Videos
聚焦教育AI生成视频的概念正确性，新评估方法解决数值错误难题
arXiv:2603.03066v2 Announce Type: replace Abstract: Existing AI-generated video quality assessment (AIGVQA) methods mainly focus on global perceptual …
7
Solving math word problems
OpenAI新系统解小学数学题，准确率追平9-12岁儿童！
We’ve trained a system that solves grade school math problems with nearly twice the accuracy of a fine-tuned GPT-3 model. It solves about 90% as many …