Anurag Mishra @_anurag_mishra

navigating loss curves, searching for meaning in local minima. i-anuragmishra.github.io New York, USA Joined May 2019

Tweets

654
Followers

116
Following

566
Likes

56K

Anurag Mishra @_anurag_mishra

2 days ago

What's up with the GPU Demand this week ??? Who is taking up so much???

1 0 0 55 0

Anurag Mishra @_anurag_mishra

a week ago

This is an amazing demonstration of what verifiers are capable of. The most important line in this thread being it was the reasoning model, not GPT 5 which selected, which solution to submit.

Mostafa Rohaninejad @MostafaRohani

a week ago

This is an amazing demonstration of what verifiers are capable of. The most important line in this thread being it was the reasoning model, not GPT 5 which selected, which solution to submit.

137 451 3K 2.2M 676

Download Image

0 0 1 122 0

Anurag Mishra @_anurag_mishra

a week ago

i know you won’t see this but Happy birthday Mr Prime Minister @narendramodi

1 0 1 30 0

Had an interesting discussion with my CEO today, and it went on multiple tangents from Tachyons to continual learning... it is fascinating how discussions like these make you realise everything is physics

0 0 1 74 0

Anurag Mishra @_anurag_mishra

2 weeks ago

The bar of intellectualism is severely low in our Indian society. Just bare articulation passes as a evidence of merit.

0 0 1 54 0

Anurag Mishra @_anurag_mishra

2 weeks ago

Road Trips ❤️

0 0 1 53 0

Download Image

Anurag Mishra @_anurag_mishra

4 weeks ago

The bar is so low….

0 0 0 67 0

Anurag Mishra @_anurag_mishra

4 weeks ago

Rage baiting is a bubble.

0 0 0 61 0

Anurag Mishra @_anurag_mishra

a month ago

Sunrise >>>> Sunset.

0 0 1 70 0

Download Image

Anurag Mishra @_anurag_mishra

2 months ago

Interesting, I do agree, that infra has to do something with it. But it does seem optimal since attention is more sensitive to quantisation and MLP‘s are not that much.

Rohan Pandey @khoomeik

2 months ago

Interesting, I do agree, that infra has to do something with it. But it does seem optimal since attention is more sensitive to quantisation and MLP‘s are not that much.

13 8 351 51K 105

1 0 1 266 0

Anurag Mishra @_anurag_mishra

2 months ago

bro is forgetting the whole point….

Marques Brownlee @MKBHD

2 months ago

bro is forgetting the whole point….

5K 10K 210K 11.7M 5K

0 0 0 189 0

Anurag Mishra @_anurag_mishra

2 months ago

This tweet instantly brings back a childhood epiphany of my own. In class 7 our teacher had just written F = ma on the board when a friend who decidedly not a physics enthusiast blurred the line between concepts and asked, “Force and energy are basically the same thing,…

Ashutosh Kumar @ashu_1069

2 months ago

6 27 783 26K 64

1 0 1 240 0

Anurag Mishra @_anurag_mishra

2 months ago

Gawwwd these conference deadlines......

0 0 0 54 0

Anurag Mishra @_anurag_mishra

2 months ago

(P, Q) = inf₍γ ∈ Γ(P,Q)₎ ∬ d(x, y) dγ(x, y)

0 0 0 60 0

Anurag Mishra @_anurag_mishra

2 months ago

When L¹ meets transport theory: EMD

0 0 0 52 0

Anurag Mishra @_anurag_mishra

2 months ago

Meaning of Peers a person of the same age, status, or ability as “another specified person”. Just in case someone needs it 😂😂😂 PS: i love when people validate inside jokes

0 0 0 54 0

Anurag Mishra @_anurag_mishra

2 months ago

it’s so disappointing to see the loss of potential to FOMO among some of my peers. But that also defines the priorities of each individual. It is hence a great social filter to your inner circle.