jake

1.3K posts

jake banner
jake

jake

@stratejake

Neural Network Mechanic @ stealth Generative models for physical systems. xoogler

SF Katılım Şubat 2014
645 Takip Edilen2.5K Takipçiler
Sabitlenmiş Tweet
jake
jake@stratejake·
My top 3 traits as an engineer: Laziness: I automate/script all repetitive tasks. Forgetfulness: I document meticulously to avoid getting lost in the future. Sloppiness: I test rigorously to catch careless mistakes.
English
4
20
295
153.1K
jake
jake@stratejake·
@lukalotl @Kellblog Yeah the other stuff on Amazon like VEEKTOMX and Ntaanoo and Gxorul is way better
English
0
0
0
86
luka
luka@lukalotl·
@Kellblog I wish they had a better logo its so ugly that I hesitate to buy even tho i like them
English
12
0
20
8.5K
Dave Kellogg
Dave Kellogg@Kellblog·
Anyone want to talk about how Anker has built a great brand seemingly without doing any marketing?
Dave Kellogg tweet media
English
401
30
3.9K
352.7K
jake
jake@stratejake·
@mirandrom @cloneofsimo so much of MLE in my experience is implementing options A, B, and C in a way that they are all a special case of D, do that you have combinatorially fewer knobs to tune.
English
0
0
1
43
Andrei Mircea
Andrei Mircea@mirandrom·
@cloneofsimo the challenge with ML is high dimensionality and imo what can make theory useful boils down to dimensionality reduction, ie finding the right abstractions that are both meaningful enough to explain a system and useful enough to give you new ways to control/influence it
English
2
1
5
1.5K
Simo Ryu
Simo Ryu@cloneofsimo·
99% people in ML is either full-eng or full-theory. I do think you really need good mix from both end of the spectrum otherwise you get really skewed perspective engineer people thinks math behind training dynamics / optimizer / diffusion / RL is useless theory people is obsessed low-mfu linear attention, over-abstraction, and toy problems that disappear at larger scale
Jason Lee@jasondeanlee

Ugh no. You need some language to even define the problem you are trying to solve. Perhaps mdp is limiting/too general but you need some formalism. You should be trying to come up with the formalism.

English
21
28
569
145.4K
Mingyuan Zhou
Mingyuan Zhou@MingyuanZhou·
@stratejake @sedielem That blog was great. We ignored how flow models were motivated and instead asked: what’s the training loss, and what’s its optimal solution? This unifies flow matching variants and lets Score identity Distillation (SiD) apply whenever a pretrained teacher provides score/velocity.
English
1
0
1
49
Sander Dieleman
Sander Dieleman@sedielem·
Two recent papers (arxiv.org/abs/2510.11690, arxiv.org/abs/2511.13720) suggest that predicting x (clean) works much better than predicting eps or v (noisy) in high dimensions. Natural signals like images live on a low-dimensional manifold. Noise takes you off the manifold! (1/3)
English
18
78
579
50.6K
Mingyuan Zhou
Mingyuan Zhou@MingyuanZhou·
@sedielem @stratejake Great discussion! In diffusion distillation, it doesn’t matter if the teacher predicts ε, x, velocity, or score — their optimal solutions are linearly related. Our paper echoes many of your points: arxiv.org/abs/2509.25127
English
1
0
1
64
teflon
teflon@teflontraphouse·
@yacineMTB homie needs a new haircut but he chill
English
4
0
7
2.5K
kache
kache@yacineMTB·
this guy is launching 12 startups in 6 months and basically you're poor and retarded 🫵😂😭
kache tweet media
English
116
32
2.1K
240.5K
xlr8harder
xlr8harder@xlr8harder·
@beffjezos @hamandcheese i don't think that fully captures it either. it tends to increase the firing rate overall, adding noise and randomness.
English
1
0
39
917
Samuel Hammond 🦉
Samuel Hammond 🦉@hamandcheese·
I did 4 cups of ayahuasca once and tunneled through my consciousness in a state of euphoria having one epiphany after another. At some point, the sense of epiphany continued without any meaningful referent, as if I was leaning on the epiphany key in my brain. That then triggered a meta-epiphany that psychedelics only make you feel like you're learning something new or deep, sorta like how déjà vu is probably just your brain's familiarity circuit misfiring in a context that isn't actually familiar. If psychedelics have given me any durable insight, it's that our consciousness is extremely fallible and often misleading, particularly with respect to valence.
Autism Capital 🧩@AutismCapital

This is probably one of the top 5 best tweets ever posted.

English
512
651
16.3K
3.4M
jake
jake@stratejake·
@tracewoodgrains Just read the "money speech" from Atlas Shrugged. That's all you really need to understand her philosophy.
English
0
0
1
50
Jack
Jack@tracewoodgrains·
thinking I should probably read an Ayn Rand book at some point to figure out what all the fuss is about will I become (more) insufferable upon doing so
English
216
3
702
88.5K
jake
jake@stratejake·
bro just invented vr
jake tweet media
English
0
0
4
232
Chief Chainsaw Officer
Chief Chainsaw Officer@Hispeedlowdrug·
The nuts n bolts section of any good hardware store is a profound testament to the honesty and decency of America's blue collar and diy types. Really makes me happy to go spend 30 mins rummaging around in one of these. A humble little corner of life that remains high trust.
Chief Chainsaw Officer tweet media
English
947
1.5K
29.6K
790.5K
Will Provost
Will Provost@GremlinIndustry·
@travis4nh The occasional party car rail trip vacation would be cool. The catch is it must be cheaper than a plane. Imagine if you rented a train car with beds and bath to travel, and it parked in a lot at your destinations. No separate hotel. Then punch in the next leg of the trip.
English
3
0
7
239
jake
jake@stratejake·
@yacineMTB Start sleep training early. We did it around 4-5 months and it was an absolute game changer. First couple of months are really hard. 6-12 is when they start to gain mobility and develop a personality and it gets a lot more fun.
English
0
0
5
221
kache
kache@yacineMTB·
please leave your unsolicited tips for newborn babies under this post
English
955
20
2K
357.3K
jake
jake@stratejake·
@bilawalsidhu @KTmBoyle Some people can't handle being given even the smallest amount of authority over others.
English
1
0
2
84
Bilawal Sidhu
Bilawal Sidhu@bilawalsidhu·
@KTmBoyle Why are people like this?! Wild to me to take the time to send such a notice
English
6
0
14
10.4K
Katherine Boyle
Katherine Boyle@KTmBoyle·
My HOA has miscalculated. They don’t know how prepared I am for war.
Katherine Boyle tweet media
English
5K
4K
156.1K
16.6M
jake
jake@stratejake·
Can I make this the default?
jake tweet media
English
0
0
2
208
Nathie
Nathie@NathieVR·
So what’s currently the best VR game out there?
Nathie tweet media
English
50
4
83
13.8K
jake
jake@stratejake·
@GrantSlatton @Gok There's a spectrum of FAANG people, from the money/status chasers to the techie problem solvers. The latter are the ones you wanna hang with outside of work, but they are an endangered species.
English
0
0
1
175
Grant Slatton
Grant Slatton@GrantSlatton·
@Gok speak for yourself ¯\_(ツ)_/¯
English
1
0
8
387
Grant Slatton
Grant Slatton@GrantSlatton·
surprising to me that technology brothers don't form more private clubs like on a sr dev faang salary, chipping in $10K/yr is affordable get 20 friends doing this, rent a commercial space, make a sick clubhouse, that's even enough to employ a full time attendant me and who?
English
25
2
259
32.8K
finbarr
finbarr@finbarrtimbers·
@stratejake not with np.testing.assert_array_almost_equal. I was surprised too!
English
1
0
1
31
finbarr
finbarr@finbarrtimbers·
I have parallel implementations of various llms in torch and jax, and I wrote tests to make them match they passed on the first try which was, uh, sus so I printed the tensors 🥲
finbarr tweet media
English
41
16
815
46.9K