思想与社会2023 · 分享于 Hacker News
GPT-3 outperforms humans on Raven's matrices, shows strong analogical reasoning
探索《GPT-3 outperforms humans on Raven's matrices, shows strong analogical reasoning》的原文与 Hacker News 社区讨论，查看来源并加入阅读清单。
从论文，走向讨论
看看社区如何讨论这个想法。
来自 Hacker News 的原始讨论。不同的视角、疑问，以及值得继续思考的线索。
最多展示 100 条相关讨论。论文可能出现在评论中，并非原帖投稿内容。
在讨论中发现的引用
按 HN 条目顺序展示最近的 50 条引用来源。
YFergusArgyll ·
Top AIs still fail IQ tests
49 赞同118 评论
Ydavidbarker ·
Governance of Superintelligence
46 赞同39 评论
Ytim_sw ·
Eight Things to Know About Large Language Models [pdf]
32 赞同7 评论
Ymarshblocker ·
Just Calm Down About GPT-4 Already
18 赞同5 评论
Yhhs ·
LLMs are not greater than the sum of their parts: Stanford researchers
4 赞同0 评论
Yog_kalu ·
GPT-3 outperforms humans on Raven's matrices, shows strong analogical reasoning
3 赞同0 评论
Yconvexstrictly ·
Sutskever: OpenAI board doing its mission to build AGI that benefits all
3 赞同0 评论
YClemDev2000 ·
Large Language Models: Scaling Laws and Emergent Properties
2 赞同1 评论
Ympweiher ·
Comparing Humans, GPT-4, and GPT-4V on Abstraction and Reasoning Tasks
2 赞同0 评论
Ymfiguiere ·
Emergent Analogical Reasoning in Large Language Models
1 赞同0 评论
YPaulHoule ·
Recursively Summarizing Enables Long-Term Dialogue Memory in LLMs
1 赞同0 评论
Yvagabund ·
Weak-to-Strong Generalization
1 赞同0 评论