Anthropic Claims Its New A.I. Model, Mythos, Is a Cybersecurity ‘Reckoning’

www.nytimes.com/2026/04/07/t...

Consider open-access and FOSS

You do research because everyone should benefit from the uncovered knowledge. Consider skipping that nice and shiny journal if people can't read your work.

Related to that; don't use file formats that lock you out (Notion or Matlab for example).

11/11

Don't let the scope creep in.

Ideally, have a clear research question and try to not expand the scope just because you have another sick idea that could be nice. I hate the amount of papers there are published every day, but sometimes a good narrative is worth more.

Get that idea out!

10/n

Be brave and bold.

The academic world back in the day was a very privileged group of people. Certain things and weird views still remain to this day. Don't let that stop bringing your views and thoughts into this field and being the change.

9/n

Keep your tools simple.

If you have any note-taking system, or something you manage your references with (which you should), make sure it gets out of your way. You don't want to fight the tools and lose your thoughts by dealing with an overly complicated setup.

Napkin notes rock. FOSS ftw.

8/n

Honest communication.

It is always easy to overestimate your capability of finishing this one (final) task. Be open about the problems you encounter. Communicate them in advance and with a plan on how you want to deal with them.

Don't oversell and underdeliver.
Guilty as charged.

7/n

Don't let LLMs edit anything on your paper.

Seriously, I was using it to fix some compilation errors uploading to arXiv and suddenly an unknown reference popped up. I did run no git diff before uploading and the rest is history. Just write your own stuff and ask your colleagues for help.

6/n

Read more.

No idea is built from scratch. The more you read, the more creative you will be approaching a problem. Better to read before you start a project and narrow the narrative to the notes you took throughout your reading sessions.

5/n

Hone your academic hard-skills.

An idea that is not written down and published (in any form) does not exist. Writing is the essential skill that everybody needs to hone.

I struggle with it, but I am trying to improve this every day. It's especially hard if you target two groups of readers.

4/n

You are studying to become a scholar.

Sometimes it can feel like you should already know your stuff when you start, otherwise, it is easy to feel like a fraud. You don't need to know how science works, you will learn it and eventually become good at it. Also, you got hired in the first place.

3/n

Take conscious breaks.

Being in a PhD is just as with the studies before, you are never really off work if you don't want to. I figured for myself that working on a problem for an extended period of time can make your life miserable if you don't step away from it and rest.

2/n

Today's my first day of the second year of my doctoral degree and I want to share some things I struggle with and things I learned throughout the first year. Just a brain dump and no particular order. Feel free to add yours :)

1/n

Offline rendering tales.

Just implemented a BVH in my own renderer using the surface area heuristic and iterative traversal.

Meanwhile, brute force still hasn't finished (30+ mins).

Been trying to implement OpenPBR (academysoftwarefoundation.github.io/OpenPBR/) in Mitsuba. Here are some preliminary results with the spectral renderer.

The geometry is the standard shader ball available in
@openusd.bsky.social. It's pretty neat to visualise different light-surface interactions.

Will Arnett Rallies Against AI in Impassioned Animation Defense From Oscars Stage Will Arnett took a stand against the use of AI during the presentation of the Academy Awards for animation.

www.thewrap.com/creative-con...

word.

SoftJAX & SoftTorch: Empowering Automatic Differentiation Libraries with Informative Gradients Automatic differentiation (AD) frameworks such as JAX and PyTorch have enabled gradient-based optimization for a wide range of scientific fields. Yet, many "hard" primitives in these libraries such as...

Found something potentially useful for the neural graphics folks: arxiv.org/abs/2603.08824

TL;DR: Package finds soft surrogates for otherwise non-differentiable functions.

Has an implementation for Torch: github.com/a-paulus/sof...

& Jax: github.com/a-paulus/sof...

Why I may ‘hire’ AI instead of a graduate student “It can competently perform a lot of the work I need immediately,” this professor writes

“In our culture, preferring an algorithm to a trainee feels like a betrayal of the academic mission.”

That’s because it is.

www.science.org/content/arti...

GitHub - NVIDIA-RTX/godot: NVIDIA fork of Godot Engine – Multi-platform 2D and 3D game engine NVIDIA fork of Godot Engine – Multi-platform 2D and 3D game engine - NVIDIA-RTX/godot

@godotengine.org received the infamous Nvidia RTX branch: github.com/NVIDIA-RTX/g...

A path tracer is included among other things.

GitHub - adobe/openpbr-bsdf: Adobe's reference implementation of the OpenPBR BSDF Adobe's reference implementation of the OpenPBR BSDF - adobe/openpbr-bsdf

Adobe open sourced their OpenPBR implementation:

github.com/adobe/openpb...

Final Fantasy XVI 14/10.

Onto the expansions!

Ray Tracing using HIP In this technical report, we introduce the basics of ray tracing and explain how to accelerate the computation of the rendering algorithm in HIP. We also show how to use a HIP ray tracing framework - ...

Ray Tracing using HIP
arxiv.org/abs/2603.00292

Is there anyone by chance who uses colemap-dh on a split columnar keyboard and has a bulletproof layout for symbols?

I still struggle to find a comfortable symbol layer.

Also found something else, which is now included in v0.2.2. I forgot to allow the sensor to yield a film plugin.

GitHub - pixelsandpointers/mitsuba-scene-description: An automatically generated Pythonic API for Mitsuba plugins to build your scenes programmatically. An automatically generated Pythonic API for Mitsuba plugins to build your scenes programmatically. - pixelsandpointers/mitsuba-scene-description

New bugfix is available for: github.com/pixelsandpoi...

Now nested plugins serialize/resolve correctly in your scene, so you can start using BlendedBSDFs :)

Are you talking about Johannes’ jo.dreggn.org/home/2015_mn...?

Fly through of part of a Keeper level directly in the Unreal Engine editor. Part way through I turn on all of the game objects the player doesn't see -- it takes a lot to make games work.

GitHub - rlguy/Blender-FLIP-Fluids: The FLIP Fluids addon is a tool that helps you set up, run, and render high quality liquid fluid effects all within Blender, the free and open source 3D creation su... The FLIP Fluids addon is a tool that helps you set up, run, and render high quality liquid fluid effects all within Blender, the free and open source 3D creation suite. - rlguy/Blender-FLIP-Fluids

I was today's years old: github.com/rlguy/Blende...

A couple of great Slang Shader talks are ready for viewing:

Vulkan releases game engine tutorial

The Vulkan Working Group has published, Building a Simple Game Engine, a new in-depth tutorial for developers ready to move beyond the basics and into professional-grade engine development.

Learn more: www.khronos.org/blog/new-vul...
#vulkan #tutorial #programming #gpu #gameengine