Pedro Castro

🇧🇷

[Pedro Castro @piotrcastro]

@fatosdabolsa Aula

[Pedro Castro @piotrcastro]

@HanpanBR E treinar mais com amiguinhos 😂

[Pedro Castro @piotrcastro]

@guidoarts_ @NotPedro3 @HanpanBR kkk

[Pedro Castro @piotrcastro]

@citadelMKT @fatosdabolsa Por favor, o senhor poderia, por gentileza, me passar esse estudo que demonstra que as ia diminuem produtividade geral das empresas. Quero muito ler sobre isso. Muito obrigado

[Pedro Castro @piotrcastro]

@HanpanBR Mano, que p* é essa ?

[basvanopheusden @basvanopheusden]

A few weeks ago, I started a new job at @OpenAI. I wrote a document about my interview process and recommendations for anyone on the job market for AI research positions. I hope it's helpful!

https://t.co/0I6f6UrAqD

[Pedro Castro @piotrcastro]

@HanpanBR @Bigbird_fgc Everybody wants Makoto, man

[Math Cafe @Riazi_Cafe_en]

MIT's Advanced Data Structures
by Prof. Erik Demaine

Lecture notes: https://t.co/EBeFGVmouo

[HanpanBR @HanpanBR]

Pra quem tiver curiosidade de como é jogar Elden Ring Night Reign, fiz um videozinho de momentos jogando com amigos.

Link no comentário.

Spoiler: há 3 partes no video: jogo normal, quando alguém fica bravo, e quando vem os trofeuzinhos.

[Santiago @svpino]

Good datasets need three things:

1. They should contain predictive features
2. They should have high-quality, diverse samples
3. They should have enough samples

You will never get away with only 2 out of these 3.

[Romee Panchal @romitheguru]

Testing your skills is one of the best ways to gauge your command of a subject. I created this Colab notebook for beginners to test their PyTorch skills and learn along the way.

https://t.co/CurVKBR8dl

[clirspec @clirspec]

📢Places are still available for the joint CLIRSPEC/CLIRPath-AI Summer School in July.
Book now to reserve your place!

https://t.co/C0qEXikEOh

@CLIRPath_AI have bursaries available for UK-based postdocs.

Follow us on Bluesky too https://t.co/ekbrK6GSgb

[Ani @curiousZeedX]

For anyone wanting to get started with Reinforcement Learning, this is probably the best resource by @OpenAI

This was originally written in 2018 but have been updated quite a few times post that.

A good starting point before you decide to dive into actual RL books/courses.

[Sichu Lu @lu_sichu]

https://t.co/2RKtpyk0mn

[Michael Friendly @datavisfriendly.bsky.social @datavisFriendly]

Dynamite plots (bars + error bars) are often the worst choice. And, there are a number of ways principalled to handle the problem of overlapping CIs.

Check out my lecture notes, https://t.co/De7H3PVp3c
and the VizTest #rstats pkg, https://t.co/X2W9v5ORdg

[Physics In History @PhysInHistory]

Interesting ways to teach physics ✍️

[Itai Yanai @ItaiYanai]

Failure is always in the background of any creative project; when the experiments are not working or the results don’t seem to make sense. But for discovery, you can’t deny yourself the permission to fail, and we should better teach failure in all its glory.

[The Nobel Prize @NobelPrize]

"Don't work too hard!" says Paul Nurse.

We especially hope that the medicine laureate isn't working too hard today as it is his birthday.

Nurse received the 2001 #NobelPrize in Physiology or Medicine for his work related to discoveries of key regulators of the cell cycle.

[clirspec @clirspec]

Prediction of post-treatment recurrence in early-stage breast cancer using deep-learning with mid-infrared chemical histopathological imaging.
npj Precis. Onc. 9, 18 (2025).
https://t.co/0g69pelc18
#infrared #breastcancer #cancer #deeplearning #ai #openaccess