0n10n

BS @PKU1898 → MSE @penn | CTO @FingerDanceSG | DM Open
Vibe Retard | Speaks Classical Chinese to save tokens

[0n10n @0xN10N]

https://t.co/6CqFbtdetI

[0n10n @0xN10N]

奥对了，也不知道 COLM 主办方怎么挑的地儿，出门就是 Union Square。然后同一条街上一边是在打药的，一边是端着电脑干活的清澈的大学生，也是很神奇了。

[0n10n @0xN10N]

去 COLM 的会场转了一圈，现在的大学生确实有点难了。大规模的训练不是谁都能做的，但调着 API 做点 Harness 比上一代 opus 啥的在某个 benchmark 上多个几个点，实在是想不明白到底有啥意义。

ICLR 已经全是泡沫了，很快这些顶会都成了抽水马桶，大家尿得都是泡沫，等着御三家来冲，然后周而复始。

[0n10n @0xN10N]

拍摄于OPENAI office门前

[0n10n @0xN10N]

@otzgary 老师你也来

[0n10n @0xN10N]

gm from San Francisco

[0n10n @0xN10N]

@0xTyche 会有的

[0n10n @0xN10N]

老马的这个robotaxi也不灵啊😅这头一回打就给我撂半路了可还行

[0n10n @0xN10N]

美元贬值了说是

[Tibo @thsottiaux]

Hi,

Tomorrow we are re-opening the Pro $200 subscriptions to new subscribers, but together with it we are also changing how we calculate the usage for it. In effect, if you do the math, it will net out at half the dollar in API spend compared to the old Pro $200 plan.

Now that it's said, let me explain why this is happening and why you will still get more work done than if you were on the Pro $200 subscription one month ago.

(a) We didn't want to compromise in other ways and are committing to not reintroducing the 5h limit, so that you can fully use the weekly usage when you want.
(b) On the subscription, we guarantee that over time you always get more work done and with an increasing level of quality. This means that you will continue to get more value per dollar spent as a result of models getting more efficient and us passing down the improvements in the form of API price reductions.
(c) We don't want to put an incentive on ourselves to artificially inflate the API list prices to make it look like you are getting a lot (and workaround it through discounts, etc). Instead we want to continue to both rapidly reduce prices and increase capabilities of models on the API. This week we introduced GPT-6 Sol and GPT-6 Luna at 50% of their previous price. Over time, we see prices go low enough that it makes sense for most to buy usage as needed without there being a significant gap between what you get in a subscription and what you get in the API for a dollar spent.
(d) Tomorrow, we are adding more things to the subscription that won't draw on the usage, I won't reveal what that is yet.

I wanted to be transparent before all the big announcements tomorrow. Lots of new exciting things are coming to the subscriptions that will make it super compelling, but I wanted to make sure to share this change ahead of time so you can all understand it before we shower you with good news.

Codexingly,
Tibo

[0n10n @0xN10N]

感谢 Mercy 老师@Mercy_okx 投喂的中秋礼盒。今年没送月饼送的是速干套装，该续个卡去健身房了

[0n10n @0xN10N]

我codex的闲置算力也跑了两周这个问题，没想到真有人搞出来了

[Alex Kontorovich @AlexKontorovich]

WOW!! Zeta(5) is irrational! Here's a Lean formalization: https://t.co/QtqzxUu0fa Amazing what we'll learn (with AI help)

[0n10n @0xN10N]

@alexandr_wang looks like intestine though

[0n10n @0xN10N]

恭喜ICLR在10年获得了100倍的增长

[Yann LeCun @ylecun]

Looks like we have over 500 submissions to ICLR this year.

https://t.co/jiawnKtnIF

[0n10n @0xN10N]

开源地址：https://t.co/i49s55gbo6

[0n10n @0xN10N]

给老师的端侧 AI 新成果宣传宣传：Qwen 27B，在16G 显卡上以256K 上下文跑到40TPS，并且效果也很 ok。

https://t.co/WcMuIST87K

[0n10n @0xN10N]

lol

[Jj McMc @AccountMus629]

@flowersslop Jev is all you need

[0n10n @0xN10N]

试玩了一小阵。很有意思的产品

传统(?)的 LLM 是任意输入，任意输出的，但是在很多工作闭环中的 agent 却只是需要其做有限选择的。

Jev 是任意输入，有限可定义的输出。这么大规模的 usecase 本来就应该是便宜+迅速的。

现在好像还不支持输入图片，但是应该不是本质的难题。

值得期待！

[0n10n @0xN10N]

搞到了Jev 的试用，感觉能整点很有意思的东西

[0n10n @0xN10N]

搞到了Jev 的试用，感觉能整点很有意思的东西

[0n10n @0xN10N]

@cyodyssey GPT 的速度慢到受不了了实在，ds 配上 dsh 的效果也不错的

[0n10n @0xN10N]

不过这个 idea 确实有意思的。之前的 AI 证明未解之谜主要是大家为爱发电，或者大公司的 PR 需求。现在孙割把它变成了一个烧 token 挖矿的过程。

[0n10n @0xN10N]

这绝逼是 codex 给写的

[H.E. Justin Sun 👨‍🚀 🌞 @justinsuntron]

https://t.co/z5PCEdqltE