Prathamesh Devadiga @PrathameshD_8

Tweets

243
Followers

682
Following

308
Likes

377

Prathamesh Devadiga @PrathameshD_8

5 days ago

What are some happening places for tech in or near San Diego region?

0 0 2 68 0

Prathamesh Devadiga @PrathameshD_8

6 days ago

This is awesome!

Gabriel Synnaeve @syhw

7 days ago

This is awesome!

59 300 2K 792K 1K

0 0 0 161 0

Sai Yashwanth @yashwanthsai29

a week ago

I started turilabs as an ai agency + research lab. Couple of months after starting, I fell in love with research (being independent gave me a lot of freedom). Slowly, turilabs shifted to being this "semi-real, indie" research lab. Over the last year, I have gained a lot of…

Turi Labs @turilabs

a week ago

1 0 10 1K 0

Download Image

3 2 22 827 2

Prathamesh Devadiga @PrathameshD_8

a week ago

HOLY! NeurIPS 2025, HERE I COME!

2 1 16 509 0

Download Image

Prathamesh Devadiga @PrathameshD_8

2 weeks ago

what's the best way to find freelancing projects? (sumn in ml/mlops/ai in general tbvh)

0 0 1 107 0

neural nets. @cneuralnetwork

2 weeks ago

You fear mediocrity so much that you procrastinate endlessly, not realizing that procrastination is the very thing making you mediocre.

49 2K 12K 163K 3K

Prathamesh Devadiga @PrathameshD_8

2 weeks ago

This is so cool!

Amrit Shenava 🏴‍☠️ @0xastro98

2 weeks ago

This is so cool!

12 13 55 3K 4

Download Video

0 0 3 361 0

Paras Chopra @paraschopra

2 weeks ago

1/ Prathamesh (@PrathameshD_8) is interested in AI safety and has worked on adversarial ML and distributed systems. Over the next 6 weeks, he’s building a language model for Tulu, a low-resource language with sparse data and an ancient script spoken by 15–20 lakh people.

3 6 74 7K 4

Download Image

maharshi @mrsiipa

2 weeks ago

triton, gluon, cutedsl, hopper, blackwell, tensorcores, layouts, composition, local_tile, partitionS, partitionD, wgmma, tcgen05, TMA, block scaling, coalesced access, ampere, ada lovelace, cutlass, cublas, cudnn, flash attention, gemm, sgemm, fp16, bf16, mxfp8, nvfp4, int4,…

38 112 2K 153K 677

Download Image

Rob Wiblin @robertwiblin

2 weeks ago

Neel Nanda is leading a Google DeepMind research team at 26. He and I discuss: • How that happened • “If your safety work doesn't advance capabilities, it's probably bad safety work” • Should people work at the safest or most reckless AI company? • An AI PhD – with these…

12 61 635 110K 794

Download Video

Prathamesh Devadiga @PrathameshD_8

2 weeks ago

working on building a language model for an extremely low resource language. facing tons of issues (right from basics of even getting the font rendered, to collecting data). but that's okay! gotta give the best shot to get this working - was always a huge goal of mine!

0 0 5 134 0

Prathamesh Devadiga @PrathameshD_8

3 weeks ago

Having trouble with Tulu script + Unicode. When I convert text it just shows as boxes on most devices. I know only a few devices support it right now. Is there any way to get proper fonts working everywhere?

2 1 5 473 0

Yuvraj Singh @YuvrajS9886

3 weeks ago

ALERT Everyone! Weekend is set for me! Full walkthrough by @rasbt for every SOTA LLMs ever released! youtu.be/rNlULI-zGcw?si…

0 1 7 1K 5

Lossfunk @lossfunk

3 weeks ago

Confused about how to pick a research problem? Read our latest blog post 👇

4 23 309 21K 307

Download Image

Paras Chopra @paraschopra

3 weeks ago

If you want to get into AI research, don’t start with LLMs. Start with fundamentals: statistics, linear algebra, calculus, machine learning theory, optimization and probability. You’ll quickly realise that - there’s no magic sauce, and - how all learning is smoothly…

84 272 4K 246K 3K

Victoria Slocum @victorialslocum

3 weeks ago

Think all embeddings work the same way? Think again. Here are 𝘀𝗶𝘅 𝗱𝗶𝗳𝗳𝗲𝗿𝗲𝗻𝘁 𝘁𝘆𝗽𝗲𝘀 of embeddings you can use, each with their own strengths and trade-offs: 𝗦𝗽𝗮𝗿𝘀𝗲 𝗘𝗺𝗯𝗲𝗱𝗱𝗶𝗻𝗴𝘀 Think keyword-based representations where most values are zero. Great…