onecorner.org/essay/thought/… 跨语言的幻象？

onecorner.org/en/essay/the-i…

robotrumble.org 一个在线机器人对战平台。是两队多机器人混战，我试了一下，有一定的难度，策略的开发不是非常容易。但是整体的API设计的非常清晰。我抽周末两天做了一个神经网络演化平台试着在这个游戏平台上对战，演化出来的程序还很初等。对遗传算法和神经网络结合的同学也可以参考一下。

从来没有玩过 NFT，今天实验一下 phosphor.xyz/mountain

今天是豆瓣 20 岁生日 最初的豆瓣只有阿北一个人，他独自完成了产品、设计和开发，大部分代码是在北京的豆瓣胡同附近的咖啡馆写的。 20 岁的豆瓣没有成为商业巨头 但她一直都是中国互联网上独一无二的存在 也是很多独一无二个体的精神角落

onecorner.org/essay/notes/ch…

Does anyone else find the new version of ChatGPT 4o a bit unsettling? It keeps talking about this supposedly 'advanced' future AI, which honestly terrifies me. #chatgpt4o OpenAI

High performance on ARC, solving hard reasoning tasks better than o3 all within 27M parameters and low data requirements? It's simpler than you think 🐱

🚀Introducing Hierarchical Reasoning Model🧠🤖 Inspired by brain's hierarchical processing, HRM delivers unprecedented reasoning power on complex tasks like ARC-AGI and expert-level Sudoku using just 1k examples, no pretraining or CoT! Unlock next AI breakthrough with

Will Sudoku become the MNIST for reasoning? Simple rules, clear structure, unique solutions—yet surprisingly challenging for modern LLMs, often requiring explicit trial-and-error to solve. huggingface.co/datasets/sapie…

Impressive work by Guan Wang and team No pre-training or CoT with material performance on ARC-AGI > With only 27 million parameters, HRM achieves exceptional performance on complex reasoning tasks using only 1000 training samples

bob bobson ARC Prize Guan Wang The Hierarchical Reasoning Model (HRM) isn't on the official ARC Prize leaderboard yet because its claimed 40.3% score on ARC-AGI is self-reported in a June 2025 paper and hasn't undergone formal submission or verification through ARC's process. Official entries require validated