jaisidhsingh Profile Banner
Jaisidh Singh Profile
Jaisidh Singh

@jaisidhsingh

Followers
192
Following
5K
Media
45
Statuses
649

machine teacher & other things

Joined August 2020
Don't wanna be here? Send us removal request.
@TVachaW
Vacha
3 days
A large part of meditation consists in overcoming the tendency to treat our thoughts / feelings as instructions on how we should act. In meditation, we can notice certain of thoughts / feelings that undermine their power to compel a certain action. We notice that thoughts /
@thestillpoint_
The Still Point
6 days
When awareness ceases to identify with the shifting contents of experience — thoughts, emotions, stories, memories — it realises itself as the open space in which all these arise. That space cannot be possessed. It has no boundary, no edge, no fixed centre. To abide as that is to
9
32
312
@jaisidhsingh
Jaisidh Singh
6 days
Bookmark this site now ( https://t.co/2Z45yhj2qs)
0
0
0
@DefenderOfBasic
Defender
8 days
people who are good at math look at an equation and immediately see the colors. You & I can't see the colors until we do a lot of extra work. But if you explain it in our language, we can see it, and solve it. It's not that complicated
@MatthewZ73671
Matthew Zeits
8 days
@DefenderOfBasic What? People don't naturally explode summations,products, and even integrals in their head?
106
534
6K
@jaisidhsingh
Jaisidh Singh
8 days
Jimmy Carr is someone I really enjoy watching, especially cause of the occasional gold he drops: “you can’t have an easy life and a great character”.
0
0
0
@PrimeIntellect
Prime Intellect
11 days
INTELLECT-3 is a 106B parameter Mixture-of-Experts model trained with both SFT and RL on top of the GLM 4.5 Air Base model. Both stages, including multiple ablations, were carried out on a 512-GPU H200 cluster over the course of two months.
9
33
559
@jaisidhsingh
Jaisidh Singh
14 days
Go read this
0
0
2
@_fracapuano
Francesco Capuano @ NeurIPS
16 days
go ahead and train language models without a single backward pass
6
21
264
@jaisidhsingh
Jaisidh Singh
17 days
Gemini 3 looks unreal on the LMArena leaderboard. Anyone use it long enough to share some thoughts?
1
0
1
@jaisidhsingh
Jaisidh Singh
18 days
Finally.
0
0
5
@jaisidhsingh
Jaisidh Singh
21 days
didn't get an answer to my research question after trying for the whole day non-stop but hey at least i have pretty plots now ;-;
0
1
3
@jaisidhsingh
Jaisidh Singh
24 days
Another day another incredible yet super slept on innovation by Shazeer
@kalomaze
kalomaze
25 days
> "...we present model architectures in which a MoE with up to 137 billion parameters is applied convolutionally between stacked LSTM layers..." - a Noam Shazeer paper that came out ~5-6 months before 'Attention Is All You Need'
0
0
2
@miniapeur
Mathieu
26 days
Pope Leo has announced his Math Texbooks picks: ✨“Algebraic Geometry” (1977, aut. Robin Hartshorne) ✨“Measure Theory” (1950, aut. Paul Halmos) ✨“Algebraic Topology” (2009, aut. Allen Hatcher) ✨“Partial Differential Equations” (1993, aut. Lawrence C. Evans) ✨“Probability
@ClubConcrave
Pope Crave
27 days
Pope Leo has announced his Criterion Closet picks: ✨“It’s a Wonderful Life” (1946, dir. Frank Capra) ✨“The Sound of Music” (1965, dir. Robert Wise) ✨“Ordinary People” (1980, dir. Robert Redford) ✨“Life is Beautiful” (1997, dir. Roberto Benigni)
22
124
1K
@Andercot
Andrew Côté
25 days
The material world is a tiny subset of the physical world. We occupy a vanishingly small volume in configuration space.
31
55
560
@francoisfleuret
François Fleuret
25 days
@Mishok2000 “Success is the ability to go from failure to failure without losing your enthusiasm” -- Winston Churchill
1
2
17
@gro_tsen
Gro-Tsen
27 days
“Dear mathematician, can you explain to me what mass is?” “Very simple, dear: the mass of a dynamical system is the cohomology class of the Galilean group representing lack of equivariance of the moment map on the symplectic manifold that is the phase space of the system.”
52
60
752
@sighjith
Saeejith Nair
28 days
I built https://t.co/VJ7oHdvf7s, a new interface for arXiv As we enter an era of accelerated scientific discovery, we need better tools that augment human cognition to help us keep up. Try it: visit papiers ai or swap arxiv -> papiers on any paper URL
@karpathy
Andrej Karpathy
5 months
I often rant about how 99% of attention is about to be LLM attention instead of human attention. What does a research paper look like for an LLM instead of a human? It’s definitely not a pdf. There is huge space for an extremely valuable “research app” that figures this out.
105
216
2K
@justinskycak
Justin Skycak
28 days
There’s a common myth that goes like this: “everybody has the same cognitive horsepower and learns at the same rate, but different people learn differently depending on their preferred learning style.” In reality, the exact opposite is true. Different people generally have
@ninja_maths
Alex Smith
28 days
"Everyone learns differently." This is a widespread and widely debunked neuromyth. Taken from The MA Way - Ch7 (link in replies).
29
122
1K