Daneul Kim

PhD student @SeoulNatlUni | Currently @AdobeResearch, Prev @MSFTResearch, @SonyAI_global

[Daneul Kim @carpedkm]

@thsottiaux When I update codex to 0.162.1 with npm, it throws error.
Error: Missing optional dependency @openai/codex-darwin-arm64. Reinstall Codex: npm install -g @openai/codex@latest
at findCodexExecutable

could you plz fix :(

[Daneul Kim @carpedkm]

@thsottiaux If it is done through quantized model, then I am not sure if it is good shipping.. Hope it is not “faster with lower performance”.

[Daneul Kim @carpedkm]

@lisa63williams @Muse It would be good if there is status page as OpenAI or Anthropic.

[Google AI @GoogleAI]

Announcing Gemini 4 Argon, our new frontier model.

Argon is built to sustain deep reasoning across complex, long-horizon workflows and delivers frontier performance in complex workflows across real-world software engineering, enterprise knowledge work like legal and finance, and cybersecurity defense.

We’re also expanding the model’s output token limit to an industry-leading 1M tokens.

Argon is currently rolling out to a set of trusted cyber defenders in the Fairwind Program, with broader availability as soon as possible.

[Daneul Kim @carpedkm]

@thsottiaux is it applied to the new $200 subscribers-only? or for all subscribers?

[Tibo @thsottiaux]

We will give one banked reset for every day you don't have access to Astra on your paid ChatGPT plan, starting today. Team is moving mountains to give access as fast as we can.

First one will land in ~ 3 hours. There is still time to create your account if you don't have one.

[Daneul Kim @carpedkm]

@thsottiaux whats wrong lol cannot work wo codex :(

[Tibo @thsottiaux]

To celebrate a week of efficiency and let you run 100'000 Luna threads this weekend... that's right... wait for it... I have reset usage limits for Codex and ChatGPT Work.

Enjoy.

[Dongjun Lee @dongjunlie]

Can a robot understand the nonverbal signals you give in real time — your pointing gestures, your gaze, the things you never put into words?

Meet EDITH: a framework that lets robots comprehend and act on human nonverbal signals.
https://t.co/giPBAA5w7j
🧵[1/n]

@KAIST_AI
#Robotics #HumanRobotInteraction #VLA #ProjectAria

[Ruiqi Gao @RuiqiGao]

We’re hiring research scientists & student researchers at Google DeepMind. DM or email me if you're interested!

I’ll be at NeurIPS this week. Happy to chat in person!

[Slater Stich @slaterstich]

Diffusion Without Tears is our attempt to make the score-matching + SDE interpretation of diffusion geometrically intuitive. If you're interested in our upcoming interview with @DrYangSong, I recommend reading this first! Link below.

[Michael Black @Michael_J_Black]

How do you rebut an AI-generated review? (This is not hypothetical -- I'm facing this now with ICCV reviews)

When writing a rebuttal, I try to understand my reviewer. What do they know? What don't they know? What's annoying them? What do they really care about?

But if a person submits an AI review, then I'm at a loss. Have they even read my paper? Have they read the AI review? What kind of person does this?

Let's say I carefully rebut the points the AI brought up. What will the reviewer do with this information? They will have no context with which to understand it. Will they feed it to the AI? Why?

My guess is that if they didn't bother to write a review themselves, they won't bother to read my rebuttal.

All I know about my reviewer is that they are lazy and couldn't be bothered to read my paper, which I labored over, and give me their thoughts. I think the only proper response is to just not reply to their review at all.

Or maybe I should give the review to an AI and ask it to write the rebuttal. Tit for tat.

PS: Yes, I'm certain the review is AI generated. It is obvious. Bland and wordy with lots of **emphasis**. I ran it through an AI detector and it's rated as 100% AI.

[#CVPR2026 @CVPR]

Looking for the EXACT #CVPR2025 decision release time? Don’t ask, we don’t know either 😉 Good luck!

Timelines: https://t.co/etLVUeselU

[Stanislav Fort @stanislavfort]

We discovered a surprising, training-free way to generate images: no GANs or diffusion models, but a ✨secret third thing✨! Standard models like CLIP can already create images directly, with zero training. We just needed to find the right key to unlock this ability = DAS
1/11

[titus @TitusTeatus]

generative editing timelapse with @krea_ai editor

adding a pool, chimney, clouds, expanding image, moving house & trees