V4.1 Flash 1017 通过自反馈 RL 强化从零搭建完整项目能力
精选理由
V4.1 Flash 1017 更新了，用自反馈 RL 让模型能从零做出完整游戏和 3D 项目，搜索功能被网友点赞。
V4.1 Flash 1017 采用基于自反馈的 RL 训练技术，重点提升从零构建完整项目的能力。官方称改进集中在游戏开发和 3D 建模两个方向。评论区有人认为其演示内容偏营销，但对其中的搜索功能给出了正面评价。
原文 · Teortaxes
> Through RL training techniques based on self-feedback, V4.1 Flash 1017 has improved its ability to build complete projects from scratch, particularly in areas such as game development and 3D modeling. X demo-slop optimized… oh no. the search part is great though https://t.co/0O74bC8pZw
