余天予 /zh/ 余天予 Wowchemy (https://wowchemy.com)zh-HansThu, 14 May 2026 00:00:00 +0000 /media/icon_hu0b7a4cb9992c9ac0e91bd28ffd38dd00_9727_512x512_fill_lanczos_center_3.png 余天予 /zh/ Deep Pre-Alignment for VLMs Thu, 14 May 2026 00:00:00 +0000 MiniCPM-o 4.5: Towards Real-Time Full-Duplex Omni-Modal Interaction Thu, 30 Apr 2026 00:00:00 +0000 MiniCPM-V/o 系列 /zh/project/minicpm-v-o/ Thu, 01 Jan 2026 00:00:00 +0000 /zh/project/minicpm-v-o/ <p>我是 <a href="https://github.com/OpenBMB/MiniCPM-V" target="_blank" rel="noopener">MiniCPM-V/o</a> 多模态模型系列的核心贡献者，参与后训练、评测、发布与迭代，并作为 <a href="https://github.com/OpenBMB/MiniCPM-V" target="_blank" rel="noopener">MiniCPM-V</a> 学生一作和 <a href="https://arxiv.org/abs/2509.18154" target="_blank" rel="noopener">MiniCPM-V 4.5</a> 第一作者推进相关工作。</p> <p>我的工作关注后训练 recipe，使高效多模态大模型在多模态理解、文档智能与真实部署场景中保持强能力。</p> MiniCPM-V 4.5: Cooking Efficient MLLMs via Architecture, Data, and Training Recipe Tue, 16 Sep 2025 00:00:00 +0000 Efficient GPT-4V Level Multimodal Large Language Model for Deployment on Edge Devices Tue, 01 Jul 2025 00:00:00 +0000 RLAIF-V: Open-Source AI Feedback Leads to Super GPT-4V Trustworthiness Mon, 27 May 2024 00:00:00 +0000 RLHF-V: Towards Trustworthy MLLMs via Behavior Alignment from Fine-grained Correctional Human Feedback Fri, 01 Dec 2023 00:00:00 +0000 Cross-Modal Omni Interaction Modeling for Phrase Grounding Thu, 01 Oct 2020 00:00:00 +0000