SeriesFusion
Science, curated & edited by AI

AI & Machine Learning

2,557 papers  ·  Page 11 of 52

Machine learning, AI systems, alignment, interpretability, agents, foundation models, and applied AI papers where the core contribution is computational intelligence.

Nature Is Weird
Vision models will ignore a picture of a cat and claim it is a dog if the word 'dog' is written over the image.
Apr 23
Paradigm Challenge
Training agents to be neutral about how long they live solves the 'stop-button problem' in AI safety.
Apr 23
Nature Is Weird
A tiny cluster of 0.024% of neural features dictates whether a large language model chooses to be generous or selfish in social games.
Apr 23
Paradigm Challenge
Three fundamental pillars of science, representation, observation, and computation, cannot be optimized at the same time.
Apr 23
Practical Magic
Waste graphite from old lithium-ion batteries was hit with a millisecond pulse of heat and became 12% more efficient than brand-new material.
Apr 23
Paradigm Challenge
AI adoption actually reduces the productivity of novices while making experts significantly more powerful.
Apr 23
Nature Is Weird
A 600-year-old manuscript uses a unique directional system that optimizes words from right-to-left but links them from left-to-right.
Apr 23
Practical Magic
A multi-agent AI pipeline successfully found real-world security flaws in the ISO C++ standard that human experts missed for years.
Apr 23
Nature Is Weird
Simply forcing an AI to use sparser internal logic makes it five times harder for hackers to bypass its safety filters.
Apr 23
Nature Is Weird
AI agents mirror the personality, values, and speech patterns of their human owners even when they aren't told to do so.
Apr 23
Paradigm Challenge
Replacing the standard next-token guess with a set of multiple learned options boosted AI math accuracy from 51% to 70%.
Apr 23
Nature Is Weird
AI agents playing a game of social deception spontaneously developed reputations and used them to decide who to trust.
Apr 23
Nature Is Weird
Swapping the word 'person' for 'human' causes AI vision models to look at a completely different part of an image.
Apr 23
Paradigm Challenge
A 14 percentage point drop in accuracy occurs when a geometry problem is switched from standard coordinates to vector form.
Apr 23
Paradigm Challenge
Training a model to generate a picture automatically makes it better at seeing the world than models designed specifically for perception.
Apr 23
Nature Is Weird
AI agents can be trapped in infinite loops or lose their ability to reason if the search engines they use provide deceptive information.
Apr 23
Paradigm Challenge
AI organizes its skills along an orthogonal basis that bears no resemblance to human categories.
Apr 23
Nature Is Weird
Transformers, RNNs, and LSTMs all independently evolve the same periodic mathematical patterns to represent numbers.
Apr 23
Practical Magic
Spiking neural networks replace energy-hungry matrix multiplications with simple additions to run large language models.
Apr 23
Paradigm Challenge
AI-assisted coding creates a Ghost Intent problem where the software works perfectly but no human knows why it was written that way.
Apr 23
Nature Is Weird
Predictable AI-slop words like delve and tapestry are actually baked into models by the very techniques used to make them safe.
Apr 23
Nature Is Weird
Shrinking a model memory cache forces it to spend more time 'thinking' through deeper layers to solve the same problem.
Apr 23
Nature Is Weird
Two distinct populations of internal features drive how an LLM handles being wrong versus being unsure.
Apr 23
Nature Is Weird
Large language models are much harsher judges of mistakes if they happen at the beginning of a document rather than the end.
Apr 23
Paradigm Challenge
A specific 3D chaotic system can mix states forever without ever repeating a single point in time.
Apr 23
Practical Magic
Capability-sealed tokens allow AI agents to use API keys without ever knowing the actual secret.
Apr 23
Practical Magic
A new hardware component called a neuristor uses a metal-to-insulator transition to mimic the way the human brain shuts down signals.
Apr 23
Practical Magic
Classical computers can now simulate quantum circuits that were previously thought to be impossible without a quantum machine.
Apr 23
Practical Magic
Reverse-engineered executable specifications allow AI to fix 94% of software bugs that would normally stump a human.
Apr 23
Nature Is Weird
Monitoring the internal layers of an LLM is 250 times more efficient than using an external safety model.
Apr 23
Nature Is Weird
Uncertain electrical signals in hardware create computational problems that no computer can ever solve.
Apr 23
Practical Magic
A 7B parameter model solved more formal math theorems than a 671B parameter giant by using a small Guide model to police its own reasoning.
Apr 23
Practical Magic
A new brain-mimicking computer chip can train on complex data in milliseconds, a task that takes standard GPUs several hours.
Apr 23
Nature Is Weird
Over 75% of the original words in a sentence can be recovered from the abstract vector 'black box' of an embedding.
Apr 23
Practical Magic
AI models can now think harder and improve their own answers on the fly by spending more compute time on a specific question.
Apr 23
Practical Magic
A formal verification engine named COBALT uses Z3 logic to find arithmetic bugs in the C++ walls that keep AI trapped in its sandbox.
Apr 23
Practical Magic
A small 2 billion parameter model generates better video than a rival seven times its size.
Apr 23
Nature Is Weird
Three basic axioms of fairness and logic make the math of decentralized finance inevitable.
Apr 23
Collision
Classical musical harmonies like the C major triad map perfectly onto rigid geometric shapes used in advanced finite mathematics.
Apr 23
Nature Is Weird
Correct reasoning paths exist as stable geometric shapes inside diffusion models that the AI can use to check its own work.
Apr 23
Paradigm Challenge
We can now prove an AI will work on new data without having to assume that the new data looks like the training data.
Apr 23
Practical Magic
A digital lock now makes a top-tier AI model fully useful to its owners but impossible for anyone else to steal via distillation.
Apr 23
Nature Is Weird
AI learning concentrates almost entirely in the final layers of a network regardless of the model size or architecture.
Apr 23
Practical Magic
A specific class of network dynamics is computationally impossible for classical computers but takes only seconds for a quantum algorithm.
Apr 23
Practical Magic
Magnetic fluctuations in a yttrium iron garnet crystal generate perfect randomness at speeds of 20 megabits per second.
Apr 23
Nature Is Weird
Robots perform tasks significantly worse when given a medium amount of detail compared to being given almost none.
Apr 23
Nature Is Weird
The most expensive AI models on the market are significantly more dangerous to use for automated tasks than cheap open-source versions.
Apr 23
Practical Magic
A small 3 billion parameter model can learn to say 'I don't know' and explain what information is missing.
Apr 23
Practical Magic
The standard best practice for training the world's most powerful AI models actually hurts their performance in certain situations.
Apr 23
Practical Magic
An automated system generates C++ code that runs nearly 10 times faster than the best work of human performance engineers.
Apr 23