X:Francois Chollet
@fchollet · X
切换来源
François Chollet@fcholletAI 评分4040
François Chollet@fcholletAI 评分2727
François Chollet@fcholletAI 评分2828
François Chollet@fcholletAI 评分2222引用Samaneh Saadat@smn_sdtJoin the Keras community meeting this Friday, October 2, 10 AM PST where we present latest developments in the Keras ecosystem. The meeting is open to anyone. Please use this link https://meet.google.com/gva-bbpr-twe to join.
François Chollet@fcholletAI 评分2626引用Samaneh Saadat@smn_sdtJoin the Keras community meeting this Friday, October 2, 10 AM PST where we present latest developments in the Keras ecosystem. The meeting is open to anyone. Please use this link https://meet.google.com/gva-bbpr-twe to join.
François Chollet@fcholletAI 评分2828
François Chollet@fcholletAI 评分2020有什么方法可以测试人工智能是否具备人类级别的通用性? “能够在新版 ARC-AGI-(n+1) 发布后立即通过它,这就是元基准测试”
引用Dylan T. Moore@dylantmoore@Jabaluck @DAcemogluMIT The meta-benchmark of being able to pass ARC-AGI-(n+1) immediately upon release. Though presumably that won't last forever.
François Chollet@fcholletAI 评分4343引用Simon Willison@simonwThe more time I spend working with coding agents, the more convinced I am that they make software engineering even harder We can do amazing things with them, but unlocking their full potential requires extraordinary discipline and knowledge
@fchollet@fcholletAI 评分2626 @fchollet@fcholletAI 评分1313 符号学习如果实现得当,在各方面都与深度学习并行、可 1:1 类比,只是表示基底不同,相应地学习机制也不同。 目标相同,一般原则相同。表示不同。但它们最终可能会收敛。
@fchollet@fcholletAI 评分2121 @fchollet@fcholletAI 评分2222 @fchollet@fcholletAI 评分4545 @fchollet@fcholletAI 评分5454
@fchollet@fcholletAI 评分2525 看来测试时扩展(test time scaling)已经出现了第三个轴:循环Transformer中的隐空间推理迭代。
引用@fchollet@fcholletTest-time scaling has two axes: running agents over longer timeframes (depth), and running a larger number of agents (breadth). Everybody knows about the first axis, but the second one is just as important when solving hard problems that require broad search.
@fchollet@fcholletAI 评分33
@fchollet@fcholletAI 评分5656 @fchollet@fcholletAI 评分4242 @fchollet@fcholletAI 评分2424 @fchollet@fcholletAI 评分3535 
@fchollet@fcholletAI 评分5959 @fchollet@fcholletAI 评分1515 @fchollet@fcholletAI 评分1313 长期以来,AI 部署的边界仅由技术能力所定义。但随着 AI 持续快速进步,它需要成为一种深思熟虑的集体选择——关于我们想要塑造怎样的世界。
@fchollet@fcholletAI 评分2121