Jong Wook Kim 💟 @_jongwook_kim Twitter profile

Pinned Tweet

Jong Wook Kim 💟

@_jongwook_kim

8 months

OpenAI is nothing without its people

10

34

594

Last Seen Profiles

@AlanGratz

@VladimrChlup

@6U7QtWa9Ip2tvo

@omertost

@wednesd71996224

@doganemreilgar

@AlexisM93572894

@pengen_stw

@stw_pdg

@YuukoFlop

@garotadobela

@Natguijarro

@NFTStudios_dev

@docencik1234

@sweetiebelleth

@mi034003

@BruceTA6

@spherefree

@Diamond18724974

@SomeOneFreeD

@Katerationopia

@superticklish26

@jandakembangstw

@BocilViral2024

@swiety_swiety

@Cr3470

@TanniehiAv

@stw_pdg

@DopamineFueled

@JingBai_ZhangBi

@SpBaystars

@Jonny_Dutch

@nF1I06LYReOhE

@AndreaTeuf97336

@jandakembangstw

@ginjoshimai

Jong Wook Kim 💟

@_jongwook_kim

8 months

♡

Sam Altman

@sama

8 months

i love the openai team so much

5K

4K

73K

8

16

380

Jong Wook Kim 💟

@_jongwook_kim

2 months

1.7x fewer tokens in Korean, which means GPT-4o feels 3.4x faster to Korean users!

21

32

267

Jong Wook Kim 💟

@_jongwook_kim

9 months

I can finally say that my voice in Whisper's ICML presentation video was AI-generated 😄

5

11

170

Jong Wook Kim 💟

@_jongwook_kim

2 years

Whisper is now available in Hugging Face transformers! Happy to learn that it's become more accessible to the 🤗 community and hoping it enables lots of amazing applications!

Arthur Zucker

@art_zucker

2 years

⏰➡👆We just released Whisper in 🤗 transformers! @openai ’s latest speech recognition transformer trained on 680 000 hours of audio! For example use case, check this notebook :

18

174

987

2

15

94

Jong Wook Kim 💟

@_jongwook_kim

4 years

@seeingwithsound @ilyasut The smallest CLIP model is available at . It’s larger than typical models for mobile, but I think it can fit and run (slowly) on mobile devices.

GitHub - openai/CLIP: CLIP (Contrastive Language-Image Pretraining), Predict the most relevant text...

CLIP (Contrastive Language-Image Pretraining), Predict the most relevant text snippet given an image - openai/CLIP

github.com

3

22

75

Jong Wook Kim 💟

@_jongwook_kim

5 years

I’m a doctor!

8

4

62

Jong Wook Kim 💟

@_jongwook_kim

5 years

MusicNet Inspector: Browse the MusicNet dataset and explore how the labels look like while enjoying the music! (feedback welcome)

0

13

34

Jong Wook Kim 💟

@_jongwook_kim

3 years

CLIP is secretly a music generation model!

Scott H. Hawley

@drscotthawley

3 years

Prompt: "a mel-spectrogram of an electric guitar playing" (VQGAN+CLIP) Won't sound great but will throw it into Griffin-Lim... You can kinda see little guitar shapes 🤣

2

14

134

0

1

21

Jong Wook Kim 💟

@_jongwook_kim

5 years

The largest GPT-2 model is now LIVE! So happy to have participated in the detection work which marks my first (co-authored) publication at @OpenAI :D Check out the new blog post and the report for our latest findings and thoughts.

OpenAI

@OpenAI

5 years

We're releasing the 1.5billion parameter GPT-2 model as part of our staged release publication strategy. - GPT-2 output detection model: - Research from partners on potential malicious uses: - More details:

58

620

2K

0

2

17

Jong Wook Kim 💟

@_jongwook_kim

2 months

GPT-4o is way better at optical character recognition in non-English / non-Latin script languages!

でょ

@Deyoyoyo

2 months

これは実用性高い左:gpt-4o 右:gpt-4

104

7K

37K

1

15

Jong Wook Kim 💟

@_jongwook_kim

4 years

Make TFRecords Usable Again!

Jong Wook Kim 💟

@_jongwook_kim

4 years

@keunwoochoi Now available with a file-level iterator that works without tensorflow :D

0

5

1

12

Jong Wook Kim 💟

@_jongwook_kim

5 years

Visualizing the MusicNet dataset for sanity check... I need to find the best way (for a machine) to learn music from these

1

4

13

Jong Wook Kim 💟

@_jongwook_kim

5 years

So excited to attend @ismir2019 and present my latest work at @nyuMARL ! Flying to the Netherlands today🚀

0

2

11

Jong Wook Kim 💟

@_jongwook_kim

3 years

Closely related to the Nature paper, my wonderful wife @nayoung_jun 's PhD research during the past few years is now on bioArXiv!

1

0

10

Jong Wook Kim 💟

@_jongwook_kim

4 years

And a shoutout to @sveltejs which made building the explorer so easier and faster!

Prafulla Dhariwal

@prafdhar

4 years

Excited to finally release what we've been working on since June! Jukebox is a small step in making neural nets produce music! Thanks to my collaborators Heewoo, @mcleavey , @_jongwook_kim , @AlecRad , @ilyasut , this work wouldn’t be possible without them!

18

44

195

0

2

8

Jong Wook Kim 💟

@_jongwook_kim

2 months

@Prashant_1722 Yes! GPT-4o has significantly improved speech recognition accuracy across languages. We don't have multilingual OCR evaluations like this yet, but you can try on the ChatGPT app!

0

6

Jong Wook Kim 💟

@_jongwook_kim

5 years

@ClementDelangue Thanks for putting up the web interface! The model is hardly perfect, but we hope this can spark further discussions in detection studies. The current model is only for GPT-2, and it’d be very interesting to build a “fits-all” detector that works with many different generators.

0

6

Jong Wook Kim 💟

@_jongwook_kim

2 months

@mynameisdjkim 아래 인용하신 분 포함 대부분 가짜입니다. 단순 개드립도 있지만 의도적으로 인기 키워드에 올라타서 클릭유도하려는 목적이 커요

1

5

Jong Wook Kim 💟

@_jongwook_kim

4 years

@keunwoochoi Now available with a file-level iterator that works without tensorflow :D

Release v0.0.7 · jongwook/tfrecord_lite

This release adds a new feature, tfrecord_lite.tf_record_iterator, which can iterate a tfrecord file. With this change, you can read tfrecord files without even tensorflow installed! Additionall...

github.com

0

5

Jong Wook Kim 💟

@_jongwook_kim

5 years

@lostanlen Vincent Lostanlen presents periodic modulation recognition in music signals

0

2

5

Jong Wook Kim 💟

@_jongwook_kim

3 years

What does retinal structure tell about the information efficiency of mammalian vision? Find out in @DukeTodayUpdate 's coverage on @nayoung_jun 's recent publications.

Living Retina Achieves Sensitivity and Efficiency Engineers Can Only Dream About | Duke Today

today.duke.edu

0

1

5

Jong Wook Kim 💟

@_jongwook_kim

5 years

Good morning #ISMIR2019 !

1

0

4

Jong Wook Kim 💟

@_jongwook_kim

2 months

@id62ai 27->23 tokens with this ChatGPT-provided translation "Halo, nama saya GPT-4o. Saya adalah jenis model bahasa baru, senang bertemu dengan Anda!". Token savings are bigger with the non-Latin alphabet languages. You can try yourself with Tiktoken !

GitHub - openai/tiktoken: tiktoken is a fast BPE tokeniser for use with OpenAI's models.

tiktoken is a fast BPE tokeniser for use with OpenAI's models. - openai/tiktoken

github.com

6

0

4

Jong Wook Kim 💟

@_jongwook_kim

2 years

Nature is beautiful!

Nayoung Jun

@nayoung_jun

2 years

Moreover, the number of cell types depends on the number of available RGCs (=channel capacity)! As we allow more neurons, the RFs are initially small in space and integrating in time; with larger numbers, RFs are larger in space and differentiating in time.

1

10

0

4

Jong Wook Kim 💟

@_jongwook_kim

3 years

@serrjoa @Dolby @ieeeICASSP @jordiponsdotme is killing it!

0

4

Jong Wook Kim 💟

@_jongwook_kim

5 years

@GretchenMarina @OpenAI @IreneSolaiman OMG I love it!!!!! I’m crying now

0

3

Jong Wook Kim 💟

@_jongwook_kim

3 years

@RiversHaveWings @pbaylies @danielrussruss I realize the legend is confusing; in that figure we meant "CLIP" as the general contrastive (language-image) pre-training method, and for that figure we still used BOW as input. Released CLIP models use transformer text encoder which can be thought of as a glorified bag of words

0

3

Jong Wook Kim 💟

@_jongwook_kim

3 years

🎊👍🏼👏🏻👩🏻‍🎓🧠🖥❤️

Nayoung Jun

@nayoung_jun

3 years

Such a precious experience, getting to meet and learn with these awesome people every day. Didn't believe it when people said you learn so much more by teaching than as a student, but now I fully admit it. Probably the 3-weeks period that I have grown THE most. Thanks @neuromatch

1

2

44

0

3

Jong Wook Kim 💟

@_jongwook_kim

5 years

Just have reset this account, hoping to keep it somewhat official, tweeting technical and research stuff primarily.

0

1

2

Jong Wook Kim 💟

@_jongwook_kim

2 years

Great walkthrough! Thanks @gordic_aleksa

Aleksa Gordić 🍿🤖

@gordic_aleksa

2 years

[🧠👨‍💻 ML Coding Series 👨‍💻🧠] Kicking off a machine learning coding series! I'll be deep-diving into the code behind many of the ML papers I've covered over the last few years - starting with @OpenAI 's CLIP! YT: @AlecRad @_jongwook_kim @ilyasut 1/

1

14

104

0

2

Jong Wook Kim 💟

@_jongwook_kim

5 years

Turns out that the MusicNet labels are not quite reliable. Track 1817 has a totally different music between 0:10 and 0:40. It also contains oboe as opposed to the label.

1

2

Jong Wook Kim 💟

@_jongwook_kim

8 months

@BorisMPower @jeremyphoward That’s how I added mine!

0

2

Jong Wook Kim 💟

@_jongwook_kim

5 years

@danagilliann Yay!!

0

1

Jong Wook Kim 💟

@_jongwook_kim

5 years

@QhansaD @dynamicBass @jordiponsdotme @sedielem @ismir2019 (psst) that’s probably a wrong Jongpil

1

0

1

Jong Wook Kim 💟

@_jongwook_kim

9 months

@heiga_zen Thank you for paving the way! 😍

1

0

1

Jong Wook Kim 💟

@_jongwook_kim

2 years

🤯

Ben Barry

@benbarry

2 years

I spent most of this past weekend experimenting with DALL·E 2, OpenAI’s new AI system that can create realistic images from a written description. I curated a book of 1000 robot paintings. You can view the entire thing online at #dalle

42

252

1K

0

1

Jong Wook Kim 💟

@_jongwook_kim

5 years

@jordiponsdotme Thank you Dr. Puig!

0

1

Jong Wook Kim 💟

@_jongwook_kim

3 years

@naotokui 今まで見たCLIPを使った生成モデルの中で一番気に入った応用です！Colabまでありがとうございます😊

1

0

1

Jong Wook Kim 💟

@_jongwook_kim

5 years

@neomoha Thank you Moha! Happy Thanksgiving!

0

1

Jong Wook Kim 💟

@_jongwook_kim

5 years

@deeplearnmusic Congratulations, Stefan!

0

1

Jong Wook Kim 💟

@_jongwook_kim

5 years

@QhansaD Thanks :D:D

0

1

Jong Wook Kim 💟

@_jongwook_kim

4 years

@seeingwithsound Forward-pass compute for the image encoder is about 4.5 GFLOPs, so it should be possible with modern mobile processors, although I have no experience on mobile AI. (I should also mention that it is intended for research and not for any deployed use cases)

1

0

1

Jong Wook Kim 💟

@_jongwook_kim

5 years

@finneco Thank you senpai!

0

1

Jong Wook Kim 💟

@_jongwook_kim

2 months

@Aizkmusic Sorry, fix on the way!

2

0

1

Jong Wook Kim 💟

@_jongwook_kim

2 months

@ihsilai Should be fixed now on the website!

Hello GPT-4o

We’re announcing GPT-4 Omni, our new flagship model which can reason across audio, vision, and text in real time.

openai.com

0

1

Jong Wook Kim 💟

@_jongwook_kim

3 years

Awesome work. Congratulations @GregDField @roy_suva @nayoung_jun @jmxpearson !

Greg D. Field

@GregDField

3 years

Latest work from our lab published in Nature today. Proud of the terrific effort and creativity of ⁦ @roy_suva ⁩ and ⁦ @nayoung_jun ⁩ and of our colab with ⁦ @jmxpearson ⁩. We show efficient coding -> coordinated tiling across cell types.

19

47

205

0

1

Jong Wook Kim 💟

@_jongwook_kim

8 months

@neomoha Hi Moha, thank you so much for your support🤍 So glad to hear from you and let’s get some beer or something when this is over

0

1

Jong Wook Kim 💟