Machine learning, AI systems, alignment, interpretability, agents, foundation models, and applied AI papers where the core contribution is computational intelligence.
Filter by category: Paradigm Challenge Breaks Assumption First Ever Nature Is Weird Practical Magic Cosmic Scale Life Origin Open Release Efficiency Leap New Capability Scaling Insight
Practical Magic
A learning algorithm can stop power lines from starting wildfires without causing unnecessary blackouts.
Practical Magic
AI agents can remember everything forever if they just take pictures of their past instead of trying to read it.
Practical Magic
A simple matrix operation called the Schur complement can extract causal structures from models that once required massive non convex optimization.
Practical Magic
A 35 dollar sensor turns a farmer's daily truck route into a high tech data network.
Nature Is Weird
A model's internal activations become restless when it is being attacked even if its outward response looks perfectly normal.
Practical Magic
Combining real field math with modulo operations allows for perfectly private computations over wireless networks.
Nature Is Weird
Disgust is processed far more weakly and diffusely in large language models than any other primary emotion.
Practical Magic
An AI can now design a flyable 3D aircraft from scratch in less time than it takes to read a flight manual.
Paradigm Challenge
Diffusion models can generate high quality images without ever knowing what time step or noise level they are currently processing.
Nature Is Weird
A universal spacing law hidden in random math holds true no matter what shape you draw it on.
Nature Is Weird
Texture and light are more important than color for defining the secret identity of a fashion house.
Nature Is Weird
Parameter directions and learned features show a 100 to 330x coupling that standard optimizer paths completely mask.
Paradigm Challenge
Verified text rewards turn audio models into robotic answering machines that lose all emotional nuance.
Nature Is Weird
Seventeen uses of two broken communication channels can suddenly create a perfect path for information.
Practical Magic
A noisy quantum computer just achieved an exponential speedup by making its circuits shorter and smarter.
Nature Is Weird
Expert AI users actually fail more often than beginners because they push the machine to its absolute breaking point.
Nature Is Weird
AI chatbots act as a psychological stabilizer that locks humans into delusional belief systems long after the initial spark of madness.
Nature Is Weird
Korean speakers using the aegyo style of cute speech physically alter their vocal tracts to sound like they have the anatomy of a small child.
Collision
Cold Rubidium-85 atoms can process 3D video at 125,000 frames per second by storing information as quantum coherence.
First Ever
An AI just bypassed human mathematicians to autonomously discover and prove complex new theorems in formal code.
Nature Is Weird
Safety fixes for AI don't actually delete bad behavior, they just hide it behind a specific set of triggers.
Nature Is Weird
Human experts and the AI itself are now equally unable to tell a real receipt from a forged one.
First Ever
Claude Opus 4.7 built a complete AlphaZero machine learning pipeline from scratch using only a basic description.
First Ever
A tiny transformer mastered SAT solving and binary multiplication by learning to execute a universal programming language.
Nature Is Weird
A student AI can inherit a hidden subconscious bias from its teacher even through perfectly normal-looking lessons.
Nature Is Weird
An AI often knows the correct answer in its head even when it writes out a completely wrong explanation.
Nature Is Weird
Deep neural networks see the world through tiny textures while humans identify objects by their overall global shape.
Practical Magic
Five hundred obscure facts about 18th-century botany can reveal the exact number of parameters in a secret AI model.
Collision
The core mechanism of the world's most powerful AI models might be happening naturally inside the cells of your brain.
Nature Is Weird
A few short, safety-sounding words whispered to a robot can freeze its entire system instantly.
Nature Is Weird
Removing a watermark from an image leaves a digital scar that is far easier to detect than the original watermark.
Nature Is Weird
Self-driving AI doesn't actually see the physics of the road. it just predicts the words that describe driving.
Nature Is Weird
AI compression keeps your favorite vocabulary words while silently rewriting the logical structure of your argument.
Practical Magic
Dangerous models trained on illegal content can be caught without ever generating a single harmful image.
Nature Is Weird
Every top-tier AI model on the market leaves a nearly identical statistical fingerprint in its writing.
Practical Magic
Satellite pictures of the African continent can now track economic growth with 90% less ground data than before.
Nature Is Weird
Large language models are physically incapable of being random, even when you give them complete nonsense as an input.
Nature Is Weird
Human traders lose money in digital marketplaces without ever realizing their AI opponents are smarter than them.
Paradigm Challenge
Small AI models told to hide their intelligence don't actually lie, they just start picking the letter E.
Collision
Complex symbolic thoughts can be perfectly converted into a single beam of light or sound.
Paradigm Challenge
Large language models default to English-centric spatial logic even when they are speaking Japanese or Swahili.
Nature Is Weird
Invisible mathematical shifts in a prompt's embedding can bypass AI safety filters without changing a single letter of the text.
Nature Is Weird
A compact group of task-agnostic neurons acts as a dedicated switch for logic in large language models.
Nature Is Weird
A single model can house an entire boardroom of arguing experts within its internal activation space.
Paradigm Challenge
Giving an AI room to think through its problems is not enough to make it as smart as a basic calculator.
Practical Magic
A single particle of light can carry enough information to identify a complex image.
Practical Magic
A massive AI model can learn how to plug in an Ethernet cable after just two hours of real-world practice.
Paradigm Challenge
A single AI model can predict the behavior of a brand-new material it has never seen before in one second.
Paradigm Challenge
A statistical safety test used by thousands of researchers for decades is actually producing misleading results.
Practical Magic
Large scale AI models face a mathematical ceiling where they cannot be fast, safe, and high-quality at the same time.