DeepSeek-V3.1 正式发布：支持混合推理架构，Agent 能力显著提升 ↗
DeepSeek 将 deepseek-chat 和 deepseek-reasoner 接口升级为 DeepSeek-V3.1。新模型采用混合推理架构，同时支持思考与非思考模式。相比 R1-0528，其思考效率更高；通过 Post-Training 优化，在工具使用与智能体任务中表现大幅提升。SWE-bench Verified 得分 66.0，Mu…
AI 雷达 / 周报
2026-W40
按原始发布周整理来源摘录与阅读线索，不自动生成个人观点或趋势结论。
本周概览
- 信号
- 7
- 主题
- 5
- 来源
- 3
重点信号
先看这些信号
本周动态 / 来源信号
智能体
DeepSeek 上线实验性多模态视觉模型 V4-Flash-Vision-Exp ↗
DeepSeek API 平台今日上线实验性多模态视觉理解模型 DeepSeek-V4-Flash-Vision-Exp。该模型在纯文本能力上与正式版 DeepSeek-V4-Flash 持平，但在需要视觉理解的 Agent Benchmark 上实现大幅跃升，多模态 Agent 能力接近 Opus-4.8。用户可通过设置 model='deepsee…
From Visual Search to Movement Control: A Priority Field for Artificial Agents ↗
Human spatial attention is widely conceptualized as being guided by a priority map that integrates perceptual salience, current goals, and past experiences. Here, we extend prio…
本周动态 / 来源信号
AI 编程
DeepSeek-V4-Flash 正式版上线，Agent 能力大幅增强 ↗
DeepSeek-V4-Flash 正式版 API 已上线公测，调用方式不变。该版本 Agent 能力显著增强，在 Terminal Bench、NL2Repo 等多个基准测试中表现远超 V4-Pro-Preview。新版原生支持 Responses API 格式并适配 Codex。模型结构与尺寸与 Preview 版一致，仅重新进行了后训练。目前仅升…
本周动态 / 来源信号
具身智能
Vision-Based 6-DoF Grasp Pose Estimation for Robot Cloth Unfolding ↗
Cloth manipulation is a challenging task due to the deformable and high-dimensional nature of cloth, which leads to complex interaction dynamics and perceptual ambiguity arising…
本周动态 / 来源信号
模型
Depth-adaptive Inference of Looped Language Models via Continuous Depth Batching ↗
A main promise of looped language models is depth-adaptive inference. By looping a block of shared layers a variable number of times, the model can use less compute for "easy" t…
本周动态 / 来源信号
多模态
Does Understanding Inform Generation in Unified Multimodal Models? From Analysis to Path Forward ↗
Recent years have witnessed significant progress in Unified Multimodal Models, yet a fundamental question remains: Does understanding truly inform generation in Unified Multimod…
值得跟进
继续跟踪
- DeepSeek-V3.1 正式发布：支持混合推理架构，Agent 能力显著提升 · 待补独立核验依据
- DeepSeek-V4-Flash 正式版上线，Agent 能力大幅增强 · 待补独立核验依据
- DeepSeek 上线实验性多模态视觉模型 V4-Flash-Vision-Exp · 待补独立核验依据
- Vision-Based 6-DoF Grasp Pose Estimation for Robot Cloth Unfolding · 待补独立核验依据
- Depth-adaptive Inference of Looped Language Models via Continuous Depth Batching · 待补独立核验依据
- Does Understanding Inform Generation in Unified Multimodal Models? From Analysis to Path Forward · 待补独立核验依据
- From Visual Search to Movement Control: A Priority Field for Artificial Agents · 待补独立核验依据