JangHyun (@janghyun_k) 's Twitter Profile
JangHyun

@janghyun_k

ML PhD student @ SNU | visiting scholar @ NYU | ex-intern @ NAVER AI

ID: 1099407274261532672

linkhttps://janghyun1230.github.io/ calendar_today23-02-2019 20:34:54

33 Tweet

69 Followers

114 Following

anton (@abacaj) 's Twitter Profile Photo

Struggling to use Flan-UL2? The flan templates from google show you exactly how to structure your prompts (based on how it was trained) Here are some examples you can follow. First up is generate a news article based on the title

Struggling to use Flan-UL2? The flan templates from google show you exactly how to structure your prompts (based on how it was trained)

Here are some examples you can follow. First up is generate a news article based on the title
Harsh Makadia (@makadiaharsh) 's Twitter Profile Photo

ChatGPT has 100,000,000+ users. But 99% don't know its output formats beyond text. 11 Powerful output formats ChatGPT can produce ( with examples ) 🧵:

Andrew (@kondrich2) 's Twitter Profile Photo

Excited to share what I’ve been working on at @openai! github.com/openai/evals is a framework for evaluating OpenAI models and an open-source eval registry. We will be granting GPT-4 access to those who submit high quality evals. Looking forward to your contributions!

near (@nearcyan) 's Twitter Profile Photo

Memorizing Transformers extends language models with the ability to memorize the internal representations of past inputs via knn-lookup into an external memory of recent (k, v) pairs, improving performance with memory size of up to 262K tokens arxiv: arxiv.org/abs/2203.08913

Memorizing Transformers

extends language models with the ability to memorize the internal representations of past inputs via knn-lookup into an external memory of recent (k, v) pairs, improving performance with memory size of up to 262K tokens

arxiv: arxiv.org/abs/2203.08913
Sudharshan (@sudu_cb) 's Twitter Profile Photo

Adobe just launched the beta of Firefly, their AI image creator. And it's incredible! - text-to-image - AI 3D modelling - AI Video editing This is the first AI image generator built towards solving creator problems. Here's a breakdown.

Stephen Wolfram (@stephen_wolfram) 's Twitter Profile Photo

2.5 months later...in collaboration with @openAI , ChatGPT gets its "Wolfram superpowers"! writings.stephenwolfram.com/2023/03/chatgp…

2.5 months later...in collaboration with @openAI , ChatGPT gets its "Wolfram superpowers"!
writings.stephenwolfram.com/2023/03/chatgp…
Sam Altman (@sama) 's Twitter Profile Photo

we are starting our rollout of ChatGPT plugins. you can install plugins to help with a wide variety of tasks. we are excited to see what developers create! openai.com/blog/chatgpt-p…

Joseph Nelson (@josephofiowa) 's Twitter Profile Photo

ChatGPT Hackathon Demos w/ @trychoma, @replit: 🔩 ChatGPU - Connect to a GPU + NVIDIA SMI 🪄 PromptWizard - Access top prompts like for games 📊 MLFlow Assist - See/deploy ML Flow experiments 🇺🇸 Political Action - Send custom emails to legislators

ChatGPT Hackathon Demos w/ @trychoma, @replit:

🔩 ChatGPU - Connect to a GPU + NVIDIA SMI
🪄 PromptWizard - Access top prompts like for games
📊 MLFlow Assist - See/deploy ML Flow experiments
🇺🇸 Political Action - Send custom emails to legislators
Kyunghyun Cho (@kchonyc) 's Twitter Profile Photo

do we want to know which variables are direct causes of a target outcome, or the full dependencies among all variables? gradually i started to think that it's probably neither, since the utility of each cause is not a function of the distance to the target outcome variable but

do we want to know which variables are direct causes of a target outcome, or the full dependencies among all variables? 

gradually i started to think that it's probably neither, since the utility of each cause is not a function of the distance to the target outcome variable but
Omar Khattab (@lateinteraction) 's Twitter Profile Photo

🔗 Thoughts on Research Impact in AI. Grad students often ask: how do I do research that makes a difference in the current, crowded AI space? This is a blogpost that summarizes my perspective in six guidelines for making research impact via open-source artifacts. Link below.

🔗 Thoughts on Research Impact in AI.

Grad students often ask: how do I do research that makes a difference in the current, crowded AI space?

This is a blogpost that summarizes my perspective in six guidelines for making research impact via open-source artifacts. Link below.
JangHyun (@janghyun_k) 's Twitter Profile Photo

I ran #notebooklm with my recent "Tareted Cause Discovery" paper (arxiv.org/abs/2408.16218). The results are super impressive! It captured the main point of the paper and explained the idea with a nice metaphor.

Vinay (@vinayp10) 's Twitter Profile Photo

In 2021, this man quit Google after a fight about his AI chatbot... And vanished overnight. But 3 years later, Google hunted him down and paid $2.7 BILLION to get him back. Here's what he knows that Google can't live without: 🧵

In 2021, this man quit Google after a fight about his AI chatbot...

And vanished overnight.

But 3 years later, Google hunted him down and paid $2.7 BILLION to get him back.

Here's what he knows that Google can't live without: 🧵
Startup Archive (@startuparchive_) 's Twitter Profile Photo

Andrej Karpathy explains what makes Elon Musk unique “I don’t think people appreciate how unique [Elon’s style] is. You read about it, but you don’t understand it—it’s hard to describe.” The first principle Karpathy — who led the computer vision team of Tesla Autopilot — has

Alex Vacca (@itsalexvacca) 's Twitter Profile Photo

BREAKING: MIT just completed the first brain scan study of ChatGPT users & the results are terrifying. Turns out, AI isn't making us more productive. It's making us cognitively bankrupt. Here's what 4 months of data revealed: (hint: we've been measuring productivity all wrong)

BREAKING: MIT just completed the first brain scan study of ChatGPT users & the results are terrifying.

Turns out, AI isn't making us more productive. It's making us cognitively bankrupt.

Here's what 4 months of data revealed:

(hint: we've been measuring productivity all wrong)