[LG]《RLP: Reinforcement as a Pretraining
[LG]《RLP: Reinforcement as a Pretraining
爱生活爱珂珂的文章
![[LG]《RLP: Reinforcement as a Pretraining](http://dayu-img.uc.cn/columbus/img/oc/1002/579702e468f587a18b349fab5c4a1aec.jpg)
![早！[太阳] 早安 ](http://dayu-img.uc.cn/columbus/img/oc/1002/9b3eed7056b13bbd89b6dfacd098d799.jpg)
早！[太阳] 早安
早！[太阳] 早安
《AI Agents from First Principles》AI智能代理从
《AI Agents from First Principles》AI智能代理从
《Redis 101 : From a Beginners POV》Redis远
《Redis 101 : From a Beginners POV》Redis远
傅里叶级数和笛卡尔坐标系有什么共同点？其实它们几乎是同一个概念的两种表现形式。核
傅里叶级数和笛卡尔坐标系有什么共同点？其实它们几乎是同一个概念的两种表现形式。核
![[人人能懂] 从并行思考、结构化学习到认知解密想知道AI如何像开“诸葛亮会”一样](http://dayu-img.uc.cn/columbus/img/oc/1002/149ab563f8e118bbae9a584b8e409ad6.jpg)
[人人能懂] 从并行思考、结构化学习到认知解密想知道AI如何像开“诸葛亮会”一样
[人人能懂] 从并行思考、结构化学习到认知解密想知道AI如何像开“诸葛亮会”一样
![[LG]《Per-example gradients: a new fronti](http://dayu-img.uc.cn/columbus/img/oc/1002/0811d9c4c71f9a8384aa523d861eb406.jpg)
[LG]《Per-example gradients: a new fronti
[LG]《Per-example gradients: a new fronti
![[CL]《Verbalized Sampling: How to Mitigat](http://dayu-img.uc.cn/columbus/img/oc/1002/d9325a95edfd5569ad988ca1c4af02e1.jpg)
[CL]《Verbalized Sampling: How to Mitigat
[CL]《Verbalized Sampling: How to Mitigat
![[LG]《Why Can't Transformers Learn Multip](http://dayu-img.uc.cn/columbus/img/oc/1002/020899494750b49b26c3242e13b82f28.jpg)
[LG]《Why Can't Transformers Learn Multip
[LG]《Why Can't Transformers Learn Multip
![[LG]《Thoughtbubbles: an Unsupervised Met](http://dayu-img.uc.cn/columbus/img/oc/1002/096f203729424466074867b3dbbc608a.jpg)
[LG]《Thoughtbubbles: an Unsupervised Met
[LG]《Thoughtbubbles: an Unsupervised Met
![[LG]《Rethinking Thinking Tokens: LLMs as](http://dayu-img.uc.cn/columbus/img/oc/1002/3753f5ce012fc8b481fb2b7f4890117b.jpg)
[LG]《Rethinking Thinking Tokens: LLMs as
[LG]《Rethinking Thinking Tokens: LLMs as
![早！[太阳] 早安 ](http://dayu-img.uc.cn/columbus/img/oc/1002/1cad1694dc8852461feb9be0342e25c3.jpg)
早！[太阳] 早安
早！[太阳] 早安
《“The G in GPU is for Graphics damnit!”:
《“The G in GPU is for Graphics damnit!”:
Thinking Machines 推出 Tinker——灵活强大的语言模型微调
Thinking Machines 推出 Tinker——灵活强大的语言模型微调
![[人人能懂] 从本质创造、跨界通感到无知之智本期节目，我们将潜入AI的“思想厨房](http://dayu-img.uc.cn/columbus/img/oc/1002/221cf6b5541c1e2baac02b2ec91e5769.jpg)
[人人能懂] 从本质创造、跨界通感到无知之智本期节目，我们将潜入AI的“思想厨房
[人人能懂] 从本质创造、跨界通感到无知之智本期节目，我们将潜入AI的“思想厨房
![[CL]《TruthRL: Incentivizing Truthful LLM](http://dayu-img.uc.cn/columbus/img/oc/1002/9466c088960ce51d76eb69aa392106ae.jpg)
[CL]《TruthRL: Incentivizing Truthful LLM
[CL]《TruthRL: Incentivizing Truthful LLM
![[LG]《Towards Verified Code Reasoning by](http://dayu-img.uc.cn/columbus/img/oc/1002/e9fff643408717015ef7cec8a11c3950.jpg)
[LG]《Towards Verified Code Reasoning by
[LG]《Towards Verified Code Reasoning by
![[LG]《Learning to See Before Seeing: Demy](http://dayu-img.uc.cn/columbus/img/oc/1002/d21e346a42e0ad847a0f026c7f1c22be.jpg)
[LG]《Learning to See Before Seeing: Demy
[LG]《Learning to See Before Seeing: Demy
![[CL]《Regression Language Models for Code](http://dayu-img.uc.cn/columbus/img/oc/1002/5ff8f601e14dabd6619985edbe537107.jpg)
[CL]《Regression Language Models for Code
[CL]《Regression Language Models for Code
![[CL]《Limited Preference Data? Learning B](http://dayu-img.uc.cn/columbus/img/oc/1002/f2d3287ccce917501ac0484ea4ade6b7.jpg)
[CL]《Limited Preference Data? Learning B
[CL]《Limited Preference Data? Learning B
作者信息
