Magic
@magicailabs · San Francisco · joined 04 Apr 2022
Aligned superintelligence.
17 800Followers
0Following
36Posts total
2.8MViews on collected posts
Postingan terbaru
Mencari postingan baru…
Frontier pretraining is said to be a big-lab-only game. We don’t have 100k chips yet, so there’s only one way: algorithmic efficiency. Our new recipe matches DeepSeek V4 Pro’s pretrain using 50x less compute – that’s roughly half the FLOPs used for GPT3, or ~$0.5M on GB200. https
460.1K views · 2.1K likes
· 176 reposts · 77 replies
Open on X →
Excited to announce we’re building an Applied Team focused on post-training. Come explore what's possible with our new (and still unreleased) LTM2 models and their 100M token context window. Apply here: https://t.co/Tx0QSIM9vI
50.7K views · 115 likes
· 9 reposts · 22 replies
Open on X →
Very excited to welcome @nvidia as Magic's latest investor! With their support, we’re looking forward to scaling long context and inference-time compute.
50.7K views · 160 likes
· 7 reposts · 9 replies
Open on X →
@magicailabs I guess I’ll be the obligatory “when access” reply.
So excited to try something like this out.
23.1K views · 194 likes
· 0 reposts · 3 replies
Open on X →
With context solved, we now focus on unbounded inference-time compute as the next (and potentially last) breakthrough we believe is needed to build reliable AGI.
Imagine if you could spend $100 and 10 minutes on one task and reliably get a great pull request for an entire
52.5K views · 449 likes
· 21 reposts · 22 replies
Open on X →
Our LTM (Long Term Memory) mechanism needs >1,000x less compute and memory than Llama 3.1 405B’s attention. Llama 3.1 would need 638 H100s *per user* to store a 100M token KV cache. LTM needs a small fraction of one.
SSMs, RNNs, and RAG all exploit weaknesses in evals like https
57.1K views · 396 likes
· 27 reposts · 22 replies
Open on X →
LTM-2-Mini is our first model with a 100 million token context window. That’s 10 million lines of code, or 750 novels.
Full blog: https://t.co/oFz4A9ynVZ
Evals, efficiency, and more ↓
1.6M views · 2.7K likes
· 420 reposts · 170 replies
Open on X →
@magicailabs @natfriedman Last option seems more interesting -- but I'm not that skilled yet https://t.co/MOFvHUFDZI
750 views · 4 likes
· 0 reposts · 0 replies
Open on X →
This round was led by @natfriedman & @danielgross with participation from @CapitalG and @eladgil, and will allow us to further scale up our models.
20.7K views · 36 likes
· 2 reposts · 5 replies
Open on X →
If you want to solve very hard problems to build safe AGI on a small team with thousands of GPUs, come join us: https://t.co/xHaNwMszLA!
16.5K views · 38 likes
· 5 reposts · 4 replies
Open on X →
We've raised $117M from @natfriedman and others to build an AI software engineer.
Code generation is both a product and a path to AGI, requiring new algorithms, lots of CUDA, frontier-scale training, RL, and a new UI.
We are hiring! https://t.co/I3fOLmrKA8
460.7K views · 671 likes
· 82 reposts · 44 replies
Open on X →
Dibandingkan akun berukuran sama
1 postingan dari 90 hari terakhir, dibandingkan dengan rentang 10K–100K pengikut. tampil luas, tetapi sedikit penonton yang merespons.
Median tayangan460 084akun ini924median untuk 10K–100K
Jangkauan, %2584.74%akun ini3.62%median untuk 10K–100K
Interaksi, %0.54%akun ini1.52%median untuk 10K–100K
| Metrik | Akun ini | Median untuk 10K–100K | Rasio |
|---|---|---|---|
| Median tayangan per postingan | 460 084 | 924 | 498× |
| Jangkauan (tayangan ÷ pengikut) | 25.8× audience | 3.62% | 714× |
| Tingkat interaksi | 0.54% | 1.52% | 0.35× |
Akun lain pada rentang ini → Bandingkan dengan akun lain → Bagaimana tolok ukur ini disusun →
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
460.7K15 Feb
16.5K
20.7K
75016 Feb
1.6M29 Aug
57.1K
52.5K
23.1K
50.7K24 Sep
50.7K21 Nov
460.1K8 Sep
Last 11 collected posts, oldest on the left. The scale is logarithmic: one post can outrun the rest a hundred times over.
Engagement rate per post
0.18%15 Feb
0.29%
0.21%
0.53%16 Feb
0.23%29 Aug
0.81%
0.98%
0.85%
0.36%24 Sep
0.31%21 Nov
0.54%8 Sep
Reactions — likes, reposts, replies and quotes — divided by views. Median for 10K–100K accounts is 1.52%.
What the audience does
Likes57.9%6 821 in total
Reposts6.4%749 in total
Replies3.2%378 in total
Quotes5.2%611 in total
Bookmarks27.4%3 229 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.
Akun serupa
LIKE 中文社区 ✓
@LIKEcoin_CN
35.6K followers
· 196 posts
Embassy of Poland U.S.
@PolishEmbassyUS
35.6K followers
· 31.5K posts
Itaú Paraguay ✓
@ItauParaguay
35.6K followers
· 20.2K posts
Eric Massaud 🍹🍹+🍹(booster) ✓
@EricMassaud
35.6K followers
· 179.3K posts
Sanoja Reinaldo #AquiNoSeLePicaTortaAEEUU
@SanojaReinaldo
35.6K followers
· 197.8K posts
🇫🇷fabholv#merciNS🇫🇷
@fabiennehol_NS
35.6K followers
· 114.3K posts
Queen Aminata クィーンアミナタ
@amisylle
35.6K followers
· 9.1K posts
Roy Rogers Happy Trails Music Shop ✓
@RoyRogers_HTMS
35.6K followers
· 49.8K posts