@sohamg121 Profile picture

Soham

@sohamg121

Research Scientist @MistralAI Previously: Google Deepmind

Joined August 2013
Similar User
Caiming Xiong photo

@CaimingXiong

Divyansh Kaushik photo

@dkaushik96

Sean (Xiang) Ren photo

@xiangrenNLP

He He photo

@hhexiy

Swabha Swayamdipta ✈️ EMNLP'24 🏖️ photo

@swabhz

Peter Henderson photo

@PeterHndrsn

Tim Vieira photo

@xtimv

Abhishek Das photo

@abhshkdz

Dani Yogatama photo

@DaniYogatama

Dipanjan Das photo

@dipanjand

Danfei Xu photo

@danfei_xu

Csaba Szepesvari photo

@CsabaSzepesvari

Marjan Ghazvininejad photo

@gh_marjan

Harsh Jhamtani photo

@harsh_jhamtani

Jonathan Lorraine photo

@jonLorraine9

Pinned

arXiv for our short paper in NAACL'19 is out: "ExCL: Extractive Clip Localization Using Natural Language Descriptions" arxiv.org/abs/1904.02755. Work with @anuva_agarwal, Zarana Parekh, Alexander Hauptmann at @LTIatCMU as part of our capstone. (1/4)


Soham Reposted

Want some TIPS? Well, then check out “Text-Image Pretraining with Spatial awareness” :) TIPS is a general-purpose image-text encoder, for off-the-shelf dense and image-level prediction. Finally image-text pretraining with spatially-aware representations! arxiv.org/abs/2410.16512

Tweet Image 1
Tweet Image 2
Tweet Image 3
Tweet Image 4

Soham Reposted

Today, we're announcing Pixtral 12B! It's our first vision model, and is really strong on the standard multimodal benchmarks, **without compromising abilities on your favourite text and reasoning tasks**. Blog: mistral.ai/news/pixtral-1… Le Chat 🐈: chat.mistral.ai/chat

Tweet Image 1

Soham Reposted

We dropped a new model - Pixtral 12B, our first-ever multimodal model. Enjoy! 🥰🎉

magnet:?xt=urn:btih:7278e625de2b1da598b23954c13933047126238a&dn=pixtral-12b-240910&tr=udp%3A%2F%2Ftracker.opentrackr.org%3A1337%2Fannounce&tr=udp%3A%2F%https://t.co/2UepcMHjvL%3A1337%2Fannounce&tr=http%3A%2F%https://t.co/NsTRgy7h8S%3A80%2Fannounce



LLMs-for-search is the new (corporate) space race.


Soham Reposted

@USAndMumbai @DrSJaishankar @DHSgov @USCIS @PMOIndia Been trying to find H1B visa appointment slot for more than 3 months, still no luck. If US has approved our H1B based on our petition, then why can’t we get the visa stamped in US?


Soham Reposted

Thread: I live in the US on a work visa. I’m among thousands of Indians unable to see their families because US consulates in India haven’t fully functioned through the pandemic. If I leave the US to see my parents, I won’t be able to return unless I get a consular appointment.


Soham Reposted

posting it here to curse your timeline

Tweet Image 1

This is the best thing on ML Twitter.

Alright folks, after a full workday of discussion with lots of nuance and zero work done, we eventually solved everything. Alexey and @MarioLucic_ suggested this solution, which I'm stealing and collecting the likes.

Tweet Image 1


Soham Reposted

Here's another awesome interactive post created by @kevinsimler that discuss some applications of complex systems in networks. meltingasphalt.com/interactive/go… You can play with simple grid-world viral transmission models that exhibit rich behavior patterns:


Soham Reposted

Maybe this is naïve–but I don't think *anyone* wants to think, at the end of their career, "gee, sure helped lots of people click on ads!"; or "sure did make cameras effective at recognizing folks!" Machine learning could be more– empowering, supporting, enabling. It should be.


This is a great loss.... 😔

It's with deep sadness that we announce that our founder and longtime director, Jaime Carbonell, has died. Dr. Carbonell was a pioneer in the field of language technologies and an inspiration to those who knew and worked with him. lti.cs.cmu.edu/news/lti-mourn…

Tweet Image 1


When you finish a PhD in computer science, they take you to a special room and explain that you must never use recursion in real life. Its only purpose is to make programming hard for undergrads.

It's with deep sadness that we announce that our founder and longtime director, Jaime Carbonell, has died. Dr. Carbonell was a pioneer in the field of language technologies and an inspiration to those who knew and worked with him. lti.cs.cmu.edu/news/lti-mourn…

Tweet Image 1


LIGHT SPEED

Pretty mind-blowing paper, 3D-printed representations of neural networks you can run inference on by shining light through... literal *light speed* inference times. They only have classification networks so far, but I'd love to see pix2pix or a GAN soon. science.sciencemag.org/content/early/…

Tweet Image 1
Tweet Image 2


Loading...

Something went wrong.


Something went wrong.