Noah Vandal
@noah_vandal
BS | MS @NDSU Interested in Neurons Building @SpeechSage Passion for biomedical arenas Born again Christian! Love a good debate, but only if there is purpose
determination > motivation every day
You don’t need a cofounder. You don’t need investors. You don’t need permission. You just need: – A laptop – Wi-Fi – And an unhealthy obsession to build something cool. 2025 is wild. People are launching million-dollar startups from coffee shops. Keep shipping. Keep showing up.…
This is a really good resource for frontend / ui design, very good model benchmarks! Check it out
New models added to the arena 🏆 Qwen3-235B-A22B-Instruct-2507 & Qwen3-Coder How will they do? Cast your vote to find out. @Alibaba_Qwen
if you are wanting to elaborate on a math concept to a non-mathy audience, i have found this is the best way
The Fourier Transform, explained in one sentence by Stuart Riffle. [bityl.co/NGqj]
New paper & surprising result. LLMs transmit traits to other models via hidden signals in data. Datasets consisting only of 3-digit numbers can transmit a love for owls, or evil tendencies. 🧵
The Era of DiffusionLM might be upon us "Diffusion Beats Autoregressive in Data-Constrained Settings" they find that DiffusionLMs outperform AR models if you’re bottlenecked by data rather than FLOPs same data can be reused for 100 epochs vs 4 epochs, as dLLM learns far more!
Open source model?
Tomorrow should be a huge day for American AI. 🇺🇸
Sometimes I think we add a barrier to learning from nasty maths. Almost as a way to gatekeep. Very sad!!!
people slowly realizing how simple stuff are in deep learning doesn’t get old
do you participate in polls without clicking the like button? (no better have the same count as likes)
this is the power of bf8 (just a guess)
Gemini 2.5 Flash-Lite is now GA — it’s our fastest (400 tokens/second), most cost-efficient ($0.10 in, $0.40 out) 2.5 model yet. Look for it in Google AI Studio + Vertex AI. 🔦😉
models are nothing without its benchmarks
Three weeks ago, we started building an AI game engine. But some models kept making things look... sloppy. So we turned finding the best one into a game. In three weeks, that game grew to 35K+ users across 135 countries. Introducing @designarena_ai, the fastest-growing…
Today, we're announcing our $32M Series A at a $300M valuation led by Insight Partners. Delve is automating busywork for humanity, starting with compliance. Here's our story.
This is who we use at @SpeechSage !! They have been awesome to work with so far
Today, we're announcing our $32M Series A at a $300M valuation led by Insight Partners. Delve is automating busywork for humanity, starting with compliance. Here's our story.
You can now run Qwen3-235B-A22B-2507 with our Dynamic 2-bit GGUFs! The full 250GB model gets reduced to just 88GB (-65% size). Achieve >5 tokens/s on 89GB unified memory or 80GB RAM + 8GB VRAM. GGUFs: huggingface.co/unsloth/Qwen3-…
Bye Qwen3-235B-A22B, hello Qwen3-235B-A22B-2507! After talking with the community and thinking it through, we decided to stop using hybrid thinking mode. Instead, we’ll train Instruct and Thinking models separately so we can get the best quality possible. Today, we’re releasing…
Turmeric vs human parasites. Curcumin is the major active ingredient of Curcuma longa L., traditionally known as turmeric and has been shown to exhibit a wide range of pharmacological activities including anti-parasitic effect. [🔬 microworlddiscove]
Energy does not just magically appear. It needs to come from somewhere. This will likely be a deciding factor on the future of the USA.
There are really 2 bitter lessons in AI. This is the second.
there is a fine line between being motivated and trying to improve yourself versus trying to keep up with the joneses. that fine line is "who are you doing it for?"