别只盯着模型参数，一个调教得当的“缰绳”竟能让Claude Opus 5在推理基准上满分通关。
Nvidia research shows that AI agents can perform well, and not go off the deep end, through fine-tuning, even if the AI model isn't that great at the …
把抽象推理变成互动游戏，用程序化世界模型攻克ARC-AGI-3，效率与智能并重。
arXiv:2607.28287v1 Announce Type: cross Abstract: ARC-AGI-3 turns abstraction into an interactive problem of skill acquisition. A player must infer an…
OpenAI两大改进让GPT-5.6 Sol在ARC-AGI-3提分188%，最严苛交互推理测试背后的关键突破。
IT之家 7 月 31 日消息，OpenAI 公司今天（7 月 31 日）发布公告，宣布通过改进框架， 让 GPT-5.6 Sol 模型的 ARC-AGI-3 提高了 188%。 IT之家注：ARC-AGI-3 由 ARC Prize 团队推出的全新交互式推理基准测试，是目前全球公认衡量 AI 通用…
两个API设置让GPT-5.6在ARC-AGI-3基准测试中得分翻三倍，推理保留与压缩是关键。
How two API settings improved GPT-5.6 performance on ARC-AGI-3, boosting scores and efficiency by retaining reasoning and enabling compaction.
低成本智能体在ARC-AGI基准上实现惊人推理性能，兼顾效率与泛化。
arXiv:2607.06764v1 Announce Type: new Abstract: Recent progress on ARC-AGI-1 from disclosed architectures has come broadly from two regimes: heavy tes…
用网格结构描述符预测符号求解器在ARC-AGI任务中的成功率，为抽象推理任务提供新视角
arXiv:2606.09026v1 Announce Type: new Abstract: We ask whether structural properties of intermediate grid states predict whether a symbolic ARC-AGI so…
提出可执行世界模型方法，在编码代理时代攻克ARC-AGI-3挑战，为通用人工智能探索新路径。
arXiv:2605.05138v2 Announce Type: replace Abstract: We evaluate an initial coding-agent system for ARC-AGI-3 in which the agent maintains an executabl…
13个曾被认为AI无法解决的ARC-AGI-2任务，被TranscendPlexity全部攻克，AI推理能力新突破。
Article URL: https://github.com/GitMonsters/13-Impossible-ARC-Tasks-SOLVED Comments URL: https://news.ycombinator.com/item?id=48240243 Points: 1 # Com…