TE
科技回声
首页24小时热榜最新最佳问答展示工作
GitHubTwitter
首页

科技回声

基于 Next.js 构建的科技新闻平台,提供全球科技新闻和讨论内容。

GitHubTwitter

首页

首页最新最佳问答展示工作

资源链接

HackerNews API原版 HackerNewsNext.js

© 2025 科技回声. 版权所有。

DeepMind Achieves Human Level Performance in Quake III Arena Capture the Flag

34 点作者 replax将近 6 年前

2 条评论

programmarchy将近 6 年前
This is pretty amazing. One thing I wondered is if this is a glorified aim bot, but it’s a “tag” version of CTF so that’s less of a factor plus they reduced the reaction time of bots. It also takes raw pixel input so it’s not gaining an advantage by having an altered game model. However, sounds like the bots didn’t receive sound input, which could be a human advantage (e.g. hearing footsteps of a non-visible player behind a wall).
dang将近 6 年前
<a href="https:&#x2F;&#x2F;news.ycombinator.com&#x2F;item?id=20054150" rel="nofollow">https:&#x2F;&#x2F;news.ycombinator.com&#x2F;item?id=20054150</a>