1 out of 500 questions in HLE(humanities last exam) by scale ai is a question with a fully wrong answer(no logical chain leads to that answer i made up fake math)
any llm that guys 500/500 is cheating :)
1 out of 500 questions in HLE(humanities last exam) by scale ai is a question with a fully wrong answer(no logical chain leads to that answer i made up fake math)
any llm that guys 500/500 is cheating :)
We just released DeepSeek-Prover V2.
- Solves nearly 90% of miniF2F problems
- Significantly improves the SoTA performance on the PutnamBench
- Achieves a non-trivial pass rate on AIME 24 & 25 problems in their formal version
Github: github.com/deepseek-ai/De…
We recently came across an interesting paper that helps LLMs be better at handling domain-specific languages like database queries or probabilistic programming languages, using an approach called "grammar prompting".
Link + brief thread below.
$64k in #bugbounty for finding basic secrets in predictable places because teams skipped Git 101 and proper .gitignore hygiene. Good on the reporter for cashing in on the perpetual lack of fundamental version control understanding.
medium.com/@sharon.brizin…
FramePack: Generate Video Forever
[NVIDIA ONLY] The creator of ControlNet just dropped a new approach to generating videos in a PROGRESSIVE way--you can generate loooong videos with LOW VRAM (just 6GB)
@SUP3RMASS1VE wrote a 1-click launcher to run the Gradio app on your PC!
FramePack: Generate Video Forever
[NVIDIA ONLY] The creator of ControlNet just dropped a new approach to generating videos in a PROGRESSIVE way--you can generate loooong videos with LOW VRAM (just 6GB)
@SUP3RMASS1VE wrote a 1-click launcher to run the Gradio app on your PC! https://t.co/AIza9iXROX
Test-Time Training (TTT) is now on Video! And not just a 5-second video. We can generate a full 1-min video!
TTT module is an RNN module that provides an explicit and efficient memory mechanism. It models the hidden state of an RNN with a machine learning model, which is updated…
Next-gen vision pre-trained models shouldn’t be short-sighted.
Humans can easily perceive 10K x 10K resolution. But today’s top vision models—like SigLIP and DINOv2—are still pre-trained at merely hundreds by hundreds of pixels, bottlenecking their real-world usage.
Today, we…
Deep Learning architectures usually aren't trained to perform search at test time, leading to sample inefficiency + poor generalization.
Latent Program Network (LPN) builds in test-time adaption by learning a latent space that can be searched. @ClementBonnet16@MattVMacfarlane
New blog post from Nvidia: LLM-generated GPU kernels showing speedups over FlexAttention and achieving 100% numerical correctness on 🌽KernelBench Level 1
This is a great infoleak exploit chain targeting YouTube by @brutecat. Love the use of a DoS flaw to make the attack stealthier!
brutecat.com/articles/leaki…
* BF16 + Stochastic Rounding doesn't always converge as well as FP32, introducing risk
* Both scaled and unscaled caution can underperform the baseline
* MARS needs more memory and compute and does not affect large-batch training
* Untuned PSGD and SOAP can lead to early…
Boring Reality LoRA just dropped for HunyuanVideo 🏙️🏞️
A fine-tune that lead not to cinematic shots, but to something that could've come out of your phone 📱
Hackathon update.
I built a programming language alongside @deepseek_ai
It's called Recursive Assembly. We emulate GPUs on CPU using finite field arithmetic.
I'm working on the docs then I'll launch tomorrow on my Substack (link in bio) #redbullfund@redbullfuturist
Hackathon update.
I built a programming language alongside @deepseek_ai
It's called Recursive Assembly. We emulate GPUs on CPU using finite field arithmetic.
I'm working on the docs then I'll launch tomorrow on my Substack (link in bio) #redbullfund@redbullfuturist https://t.co/BuoRTIWpBi
128 Followers 341 Followingproud of you | building ResumeChecker 📄and TabWarrior |https://t.co/wbBtg7pSXc - $0 MRR | Indiana University Informatics | i like hockey
1K Followers 6K FollowingRedbean is the social platform for beginner game creators. Use AI to turn your ideas into interactive games, share your creations -no coding!
@ycombinator S21
2K Followers 7K Following"Without the United States everything in the world would die". - DONNIE T.
Lore Master
MechWarrior Online Worlds 2025
Ranked 8th out of 86 teams.
12K Followers 1K FollowingSenior AI Reporter, Ars Technica. Tech Historian. Fast Company / The Atlantic / Retronauts / Creator https://t.co/Rh4KGhtWM0, The Culture of Tech
2K Followers 6K FollowingExpert Recruitment "Headhunter" in Technical, Digital Marketing, Blockchain/web3.0 & AI 🐋
Let me HELP you build a Brilliant Team - Big Wave Digital.
37 Followers 382 FollowingLet's Grow your Business Safely. With Using Our Service. Get Here All Kinds Of Real, Organic, Genuine, Legit full verified All accounts and All reviews Service
25 Followers 189 FollowingUse AI to create any character, with any art, chat with your character, and play them in any game!
Create D&D characters, MTG cards, Pokemon cards, and more.
484 Followers 314 FollowingNeuroscientist. Head of Neuromotor Interfaces, VP Research,
@Meta Reality Labs. Same @ the other app.
Helping @transalt @zuckermanbrain
7K Followers 598 FollowingHacking neural networks so that we don’t get stuck in the matrix. Builder and Breaker. Opinions are my own. https://t.co/ij8buvMaXg
131 Followers 178 FollowingAlone in the basement, building minds that won’t abandon you. Local LLMs, autonomy, ethics, persistence. The Blue Collar AI Tech you don’t know you need yet.
12K Followers 1K FollowingRaising kids & bread & grant money. Cleaning data & diapers & fish. EA (bed nets not light cone). Social scientist. https://t.co/g8teKfCf91
16K Followers 709 FollowingML Engineer @ML6team, part-time at @huggingface. @KU_Leuven grad. General interest in machine learning, deep learning. Making AI more accessible for everyone!
7K Followers 1K FollowingApplied AI/ML & Full Stack dev.
Optimizing Small Medium Enterprises with AI tooling and fundamental software.
destroyer of b2b SaaS integrations
5K Followers 4K FollowingEx-CCO@Tenstorrent. 元社外取締役@Sanrio. 元レノボ兼NEC PC代表取締役執行役員社長. Ex-VP@Lenovo, CEO@NEC PC, CVP@AMD. Ex-🇨🇦,🇸🇬,🇯🇵. Living in Austin🇺🇸なう. Opinions are my own.
87K Followers 194 FollowingBuilding beautiful things like Mojo🔥 and MAX @Modular, lifting the world of production AI/ML software into a new phase of innovation. We’re hiring! 🚀🧠
8K Followers 1K Following25 yrs using/researching/buying #supercomputers. Now an engineering leader for #supercomputing capabilities at Microsoft. Posts on #HPC #AI #cloud #F1 #travel
102K Followers 43 FollowingBuilding the Android of self-driving cars.
comma 3X is available now for $999, plugs into the car you already drive, and drives half your miles.
2K Followers 507 FollowingSenior Research Fellow @EPCCed, University of Edinburgh. Interested in novel architectures, HPC, FPGAs, RISC-V, programming language design and LLVM & MLIR.
3K Followers 204 FollowingCEO of JabPerf, Contributing Author to "Performance Analysis & Tuning on Modern CPUs" (1st & 2nd Edition available on Amazon), Blogger, and Former Amateur Boxer
543K Followers 23K FollowingThe best from ML/AI community | Ex-Microsoft, Rackspace, Fast Company | Wrote eight books about the future | Silicon Valley robots, holodecks, BCIs, & startups.