LogoThread Easy
  • 発見
  • スレッド作成
LogoThread Easy

Twitter スレッドの万能パートナー

© 2025 Thread Easy All Rights Reserved.

探索

Newest first — browse tweet threads

Keep on to blur preview images; turn off to show them clearly

迪卡侬的部分运动用品、任何地方都不发运动、有固定计划风雨无阻、买了苹果表、是

迪卡侬的部分运动用品、任何地方都不发运动、有固定计划风雨无阻、买了苹果表、是

独立科技网站 - 蓝点网 / 感谢关注 订阅频道:https://t.co/xzeoUEoPcU 联系方式:https://t.co/LJK1g3biPp

avatar for 蓝点网
蓝点网
Thu Nov 06 10:56:56
今天问了一下AI,了解了一下之前发生的事情,唏嘘不已
-------
有利网的前 CEO 是刘雁南。他于 2012 年联合吴逸然、任用共同创立有利网,并担任 CEO 至 2015 年 9 月。

在他的领导下,有利网成为软银中国在内地投资的首家互联网金融企业,并完成多轮融资,累计成交额突破 25 亿元。

离开有利网后,刘雁南于 2015 年创立美利金融,聚焦消费金融领域,旗下设立了专注二手车分期的 “美利车金融” 和消费分期的 “有用分期”。

美利金融曾获得新希望集团等机构的投资,截至 2019 年 6 月,美利车金融的汽车融资贷款余额达 219 亿元。

然而,2019 年 11 月,有用分期因涉嫌 “套路贷” 被定性为 “1105” 特大涉黑网络套路贷专案,刘雁南被警方抓获。

此外,刘雁南还曾在 2018 年参与区块链项目Engine的早期投资,该项目旨在通过区块链技术解决二手车交易中的信息不对称问题。

但受限于法律风险和行业监管,其后续动态未再有公开报道。

今天问了一下AI,了解了一下之前发生的事情,唏嘘不已 ------- 有利网的前 CEO 是刘雁南。他于 2012 年联合吴逸然、任用共同创立有利网,并担任 CEO 至 2015 年 9 月。 在他的领导下,有利网成为软银中国在内地投资的首家互联网金融企业,并完成多轮融资,累计成交额突破 25 亿元。 离开有利网后,刘雁南于 2015 年创立美利金融,聚焦消费金融领域,旗下设立了专注二手车分期的 “美利车金融” 和消费分期的 “有用分期”。 美利金融曾获得新希望集团等机构的投资,截至 2019 年 6 月,美利车金融的汽车融资贷款余额达 219 亿元。 然而,2019 年 11 月,有用分期因涉嫌 “套路贷” 被定性为 “1105” 特大涉黑网络套路贷专案,刘雁南被警方抓获。 此外,刘雁南还曾在 2018 年参与区块链项目Engine的早期投资,该项目旨在通过区块链技术解决二手车交易中的信息不对称问题。 但受限于法律风险和行业监管,其后续动态未再有公开报道。

从投资领域转到创业:找工作、找面试题、改简历、模拟面试. 创业(冷启动)|AI , AIGC | 安全技术|RAG | 时空智能 | 认知心理学|智能体 | 生命科学 | 强化学习 I built open source software at https://t.co/b69DXZhcyR

avatar for Y11
Y11
Thu Nov 06 10:54:21
Remember <marquee>?

Remember <marquee>?

Founder @Tailscan for Tailwind CSS Co-Founder @Lexboostai + many random side projects: https://t.co/TPk3m9LhZa, https://t.co/uW4shohLZq, https://t.co/BFujf7veHX

avatar for Erwin
Erwin
Thu Nov 06 10:52:57
How Kimi-K2-Thinking compares to MiniMax M2? 1/n

Both support interleaved tool calling

MiniMax M2 will give it a tough fight - it has made its own niche.

It features a context window of 196.6K tokens. The model costs $0.15 per million tokens for input and $0.45 per million tokens for output.

Comparing with below chart, MiniMax M2 is literally 75-80% cheaper compared to Kimi-K2-Thinking. 

But things become interesting when you use Kimi-K2-Thinking with caching & you have long context and multiple cache hits. Then Kimi K2 price for input tokens drops to MinMax M2 level.

How Kimi-K2-Thinking compares to MiniMax M2? 1/n Both support interleaved tool calling MiniMax M2 will give it a tough fight - it has made its own niche. It features a context window of 196.6K tokens. The model costs $0.15 per million tokens for input and $0.45 per million tokens for output. Comparing with below chart, MiniMax M2 is literally 75-80% cheaper compared to Kimi-K2-Thinking. But things become interesting when you use Kimi-K2-Thinking with caching & you have long context and multiple cache hits. Then Kimi K2 price for input tokens drops to MinMax M2 level.

How Kimi-K2-Thinking compares to MiniMax M2 on size? 2/n 1. MiniMax M2 has 10B active and 230B total parameters with full attention. 2. Kimi K2 has 35B active and 1 trillion total parameters. Both have most of their weights in 8bits. That mean M2 will be much easier to host and its KV cache will be much more compact. MiniMax M2 uses full attention, would be interesting to see if Kimi-M2 has done something interesting to the attention layer. (for this calculations I am assuming Kimi-K2-Thinking is based on Kimi-K2-Base)

avatar for GDP
GDP
Thu Nov 06 10:49:24
RT @santhonydo: Join a community of builders online! 

I met some cool ppl @deifosv @goon_nguyen @navinpeiris @poppacalypse @poyashad @jaso…

RT @santhonydo: Join a community of builders online! I met some cool ppl @deifosv @goon_nguyen @navinpeiris @poppacalypse @poyashad @jaso…

Indie hacker, solopreneur building a portfolio of products & SaaS: 🔌 https://t.co/rLuxXYrO8V 🟩 https://t.co/f93dEmKKZU 📋 https://t.co/zmLERuwToj ✍️ https://t.co/MW9HQLABxB 🏡 https://t.co/EtGIs3qLMx

avatar for Jason Leow
Jason Leow
Thu Nov 06 10:48:34
A fat poet (32 billion activated parameters and 1 trillion total parameters) with impressive tool calling abilities will be in the town.

Kimi-K2-Thinking supports Interleaved Tool calling

A fat poet (32 billion activated parameters and 1 trillion total parameters) with impressive tool calling abilities will be in the town. Kimi-K2-Thinking supports Interleaved Tool calling

AI @amazon. All views personal!

avatar for GDP
GDP
Thu Nov 06 10:46:25
  • Previous
  • 1
  • More pages
  • 734
  • 735
  • 736
  • More pages
  • 2127
  • Next