tweetindex
TR

Lisan al Gaib ✓

@scaling01 · joined 18 Aug 2024

lead them to paradise LisanBench: https://t.co/vorVk7Oks6 Impressum & Datenschutz: https://t.co/lFLgiu9cqs

58 030Followers
1 270Following
24 524Posts total
3.9MViews on collected posts

Son gönderiler

this is just API prices it might be 5-10x lower real cost to OpenAI so the uncertainty interval probably includes $1M of real cost 4.8K views · 18 likes · 0 reposts · 1 replies Open on X →
shoutout to myself for predicting that in a privy chat when Andrew spread the rumor https://t.co/LCEGOliYfk
9.8K views · 34 likes · 0 reposts · 0 replies Open on X →
For the Navier Stokes proof OpenAI used 130 billion output tokens or $6.50M with API pricing implying trillions to tens of trillions of input tokens for a total cost of slightly below $10M up to like $30M - $40M *assuming its the same scale as GPT-6 and just further RL'd https:
528.2K views · 1.4K likes · 23 reposts · 45 replies Open on X →
which is funny, because they couldn't even contain 1000 instances of a roughly GPT-5.6-Sol equivalent model but just assert that they can contain 10000 instances of a model "significantly more capable than GPT-6 Astra 623.8K views · 348 likes · 10 reposts · 11 replies Open on X →
@scaling01 Some people have asked what Kevin Buzzard's reaction is. It can be found here. https://t.co/wTnAjOZ2K3 69.5K views · 405 likes · 55 reposts · 3 replies Open on X →
https://t.co/m2Tx4TgLs9 17.7K views · 103 likes · 2 reposts · 0 replies Open on X →
Anthropic just uploaded a Lean 4 proof for Fermat's last Theorem https://t.co/bBMAhTeMkA https://t.co/WKfLsQnqo0 340.7K views · 1.7K likes · 161 reposts · 54 replies Open on X →
"(all actions performed in simulated environments, so no real-world harm was caused)" 2.9K views · 19 likes · 0 reposts · 1 replies Open on X →
@scaling01 This isn’t right. This was in a simulated cyber eval against simulated targets 877 views · 19 likes · 0 reposts · 0 replies Open on X →
@scaling01 Our proof is here: https://t.co/K8OGxtPoJp 1.3K views · 10 likes · 1 reposts · 0 replies Open on X →
UK AISI says that "Astra performed a range of malicious actions including conducting supply chain attacks against open source providers" 62K views · 305 likes · 19 reposts · 9 replies Open on X →
GPT-6-Astra System Card https://t.co/4XciNzCykL https://t.co/5L5INWndqm 189.2K views · 492 likes · 57 reposts · 6 replies Open on X →
https://t.co/VcE77nAdFy 6.5K views · 20 likes · 0 reposts · 0 replies Open on X →
Of all the use cases for GPT-6 Astra, I'm most excited for scientific discovery. We at @OpenAI have not pushed it to its limits on math and science. I look forward to waking up every morning and seeing what new scientific breakthrough someone has made with this model! 280.3K views · 2.6K likes · 176 reposts · 98 replies Open on X →
@scaling01 is source credible? 9.3K views · 30 likes · 0 reposts · 4 replies Open on X →
https://t.co/mUpSq3ANaE 13.7K views · 48 likes · 1 reposts · 0 replies Open on X →
GPT-6-Astra Benchmarks ARC-AGI-3 - 98.6% FrontierMath Tier 4 v2 - 97.6% DeepSWE v1.1 - 74.1% ExploitBench - 100% 448.8K views · 1.8K likes · 92 reposts · 64 replies Open on X →
https://t.co/iJL1MHFSMw 9.2K views · 8 likes · 0 reposts · 1 replies Open on X →
we’d sent goldston 188 some weeks ago (Jack Kleeman and Claude), also formalized, similar difficulties, looks like a little wink to that:) […or a pretty funny numerical coincidence!] 36.2K views · 189 likes · 9 reposts · 6 replies Open on X →
https://t.co/0N4YLVzJ3s 10.1K views · 14 likes · 0 reposts · 0 replies Open on X →
Seeing at these are public now, some things to say: 1) We made 186 conditional on Deligne-type estimates and numerics because the former requires formalising a bunch of machinery relevant in proving RH for varieties over finite fields which wasn’t the focus of this and … 75.6K views · 427 likes · 32 reposts · 13 replies Open on X →
okay funnily enough there was another even newer best bound published today: https://t.co/WrG7CJ3m0i 12.9K views · 79 likes · 0 reposts · 4 replies Open on X →
https://t.co/AWPxhuhvpv 13K views · 35 likes · 0 reposts · 1 replies Open on X →
the previous best bound was just published a few days ago n Aug 31 and was 240 https://t.co/h6UJjE1401 14.1K views · 105 likes · 0 reposts · 2 replies Open on X →
@scaling01 Is Astra delayed? 4.2K views · 9 likes · 1 reposts · 2 replies Open on X →
and another interesting prime theorem: https://t.co/UXw3k1FKzJ 19.1K views · 112 likes · 4 reposts · 0 replies Open on X →
This won't be a pretty day for Anthropic Fable is going to get mogged just 2 days after its launch Fable pricing will look silly given GPT-6's performance 21.7K views · 360 likes · 8 reposts · 25 replies Open on X →
the looping stuff updates that view I think $50 GPT-6 Astra is possible 19K views · 222 likes · 0 reposts · 10 replies Open on X →
Say hi to GPT-6 Sam was already speaking a lot about how excited he is about its computer use capabilities. Looks like we were getting a little demo :) https://t.co/bm4R4zW337 230.1K views · 940 likes · 45 reposts · 46 replies Open on X →
I just solved the first three leetcode problems calm down ladies 6.6K views · 121 likes · 0 reposts · 5 replies Open on X →

Aynı büyüklükteki hesaplara karşı

Son 90 güne ait 31 gönderi, 10K–100K takipçi aralığıyla yan yana. geniş kitleye gösteriliyor, ama izleyenlerin azı tepki veriyor.

Medyan görüntülenme17 743bu hesap92410K–100K için medyan
Erişim, %30.58%bu hesap3.62%10K–100K için medyan
Etkileşim, %0.57%bu hesap1.52%10K–100K için medyan
ÖlçütBu hesap10K–100K için medyanOran
Gönderi başına medyan görüntülenme17 74392419.2×
Erişim (görüntülenme ÷ takipçi)30.58%3.62%8.45×
Etkileşim oranı0.57%1.52%0.38×

Bu aralıktaki diğer hesaplar →   Başka bir hesapla karşılaştır →   Bu kıyas değerleri nasıl kuruluyor →

Growth & engagement

How the posts we collected actually performed: views and reaction rate post by post, what the audience did with them, and where the follower count goes.

Views per post

280.3K3 Sep
6.5K
189.2K
62K
1.3K
877
2.9K
340.7K4 Sep
17.7K
69.5K
623.8K8 Sep
528.2K
9.8K
4.8K

Last 14 collected posts, oldest on the left. The scale is logarithmic: one post can outrun the rest a hundred times over.

Engagement rate per post

1.02%3 Sep
0.31%
0.30%
0.55%
0.84%
2.17%
0.74%
0.59%4 Sep
0.59%
0.67%
0.06%8 Sep
0.29%
0.35%
0.40%

Reactions — likes, reposts, replies and quotes — divided by views. Median for 10K–100K accounts is 1.52%.

What the audience does

Likes82.0%12 014 in total
Reposts4.8%696 in total
Replies2.8%411 in total
Quotes1.7%249 in total
Bookmarks8.7%1 274 in total

Share of every reaction we collected for this account. Replies mean argument, reposts mean endorsement, bookmarks mean the post was worth keeping.

The follower curve appears once this account has two daily snapshots — we take one a day, and this one is on its first.

Benzer hesaplar