2420 位用户此时在线
24小时点击排行 Top 10：
- 本站自动实时分享网络热点
- 24小时实时更新
- 所有言论不代表本站态度
- 欢迎对信息踊跃评论评分
- 评分越高，信息越新，排列越靠前
2
1
0
Large Reasoning Model时代, 几乎等于Reinforcement Learning + LLM的时代。
但RL专业性非常强，去参加ML的会议时，专门做RL的研究员都现场拿着笔纸推算数学公式，掌握起来学习难度较高。
分享一本RL的入门教材，从RL基础MDP，PPO，直到跟LLM结合，如RLHF，都有讲解，深入浅出。
Reinforcement
时政
(
twitter.com
)
由
马东锡 NLP 🇸🇪
提交
Large Reasoning Model时代, 几乎等于Reinforcement Learning + LLM的时代。
但RL专业性非常强，去参加ML的会议时，专门做RL的研究员都现场拿着笔纸推算数学公式，掌握起来学习难度较高。
分享一本RL的入门教材，从RL基础MDP，PPO，直到跟LLM结合，如RLHF，都有讲解，深入浅出。
Reinforcement Learning: An Overview：
https://t.co/rjYSpOtbJl
点击图片查看原图
1
2
1
1
2
2
1
1
3
2
1
1
4
2
1
1
5
ANDREJ KARPATHY: ELON LOVES SMALL, HIGHLY TECHNICAL TEAMS.
2
1
1
ANDREJ KARPATHY: ELON LOVES SMALL, HIGHLY TECHNICAL TEAMS.
“Elon likes very small, strong, highly technical teams.
At companies, by default, the teams grow and they get large.
Elon was always a force against growth.
I would have to work and expend efforts to hire people.
I
btc
(
twitter.com)
00:00:33
6
2
1
1
7
2
1
1
8
2
1
1
9
2
1
1
10
2
1
1
11
2
1
1
12
2
1
1
13
GROK 4 SEES, TALKS, AND THINKS OUT LOUD
2
1
1
GROK 4 SEES, TALKS, AND THINKS OUT LOUD
It doesn’t just chat - it watches, analyzes, and answers in real time.
Grok 4’s voice mode is scary good now: more human, less robot, and yes - it speaks Spanish, French, Japanese, and even rambles like it’s got thoughts of its own.
You
btc
(
twitter.com)
00:00:50
14
We hit $10M ARR in 2 months making us one of the fastest growing AI startups in the world 🚀
2
1
1
We hit $10M ARR in 2 months making us one of the fastest growing AI startups in the world 🚀
And today, we are putting more power into the hands of people while changing the world of vibecoding 🙌
Introducing:
1. Emergent Pro Mode- Build real autonomous agents
2. Mobile Apps-
时政
(
twitter.com)
00:00:17
15
2
1
1
16
2
1
1
17
2
1
1
18
2
1
1
19
2
1
1
20
2
1
1
21
2
1
1
22
2
1
1
23
2
1
1
24
2
1
1