PAPER PLAINE

Seizure-detection algorithms perform well on clean training data but can fail when hospitals add noise, equipment variation, or interference—problems that don't show up in standard tests. Researchers created RobustSeiz, an open-source stress-testing tool that deliberately introduces realistic disruptions to see how robust these AI systems really are before they're deployed to patients.

Researchers created a new watermarking method for large language models that identifies AI-generated text while running 6000 times faster than existing techniques and adding less than 1% to generation speed. The method works by making a single random decision per word instead of searching through lists or running complex procedures, yet maintains the same ability to detect AI text with statistical certainty.

A new method called ESPO improves how AI systems are optimized through text instructions, achieving 3.76 percentage points higher accuracy than previous approaches while cutting prompt length by nearly half. The key innovation is breaking optimization into three distinct phases—identifying error patterns, generating diverse candidate solutions, and selecting the most stable ones—which prevents the bloat that plagued earlier methods where longer prompts didn't actually work better.

This paper solves a long-standing puzzle in trajectory optimization: how to find the mathematically optimal path when a spacecraft must satisfy conditions defined by equations rather than fixed points. The solution introduces new rules for when a trajectory satisfies these equation-based constraints, revealing that optimal paths sometimes behave in counterintuitive ways that don't match classical optimization theory.

When an AI agent can abandon its reputation and start fresh with a new identity at low cost, reputation stops working as a disciplinary force. Researchers modeled this problem mathematically and found that an agent's willingness to cheat depends on how expensive it is to reset identities, how quickly reputation fades, and how much customers care about past performance—showing that cheap identity switches can make reputation nearly useless at preventing fraud.

Researchers developed a method that converts natural-language instructions into small, reusable AI functions that run locally without needing to call expensive remote models. On difficult test cases where other approaches failed entirely, their method achieved 83.6% accuracy while producing functions that can be stored, shared, and combined like ordinary software — though generating them takes about a minute rather than seconds.

When one person's treatment affects their neighbors' outcomes — common in social networks — researchers face a puzzle: how do you estimate what would happen under a different policy than the one that generated your data? This paper solves that puzzle by constructing weights that mathematically transport data from one policy to another, even when interference patterns are misspecified, and provides tools to measure the resulting bias-variance trade-off.

A trained autoencoder's internal parameters contain a fingerprint of the data it learned from. Researchers showed that by analyzing the mathematical structure of these parameters—specifically their singular values—you can tell which datasets a model was trained on with high accuracy, without ever looking at the original data or using complicated reconstruction methods.

Language models make probabilistic forecasts that contain internal contradictions—you could set up bets against their predictions and guarantee a profit. Researchers tested this by generating stock market scenarios and finding that language models' probability estimates violate basic logical consistency, with contradictions growing worse when events are logically related and worsening dramatically when irrelevant details are added to the scenario.

Researchers discovered that AI language models like Gemma internally represent the different symptoms of depression in a way that matches how clinicians assess them. By examining the model's internal computations, they found a specific layer where mood problems, physical symptoms, and suicidality thoughts separate into distinct patterns, and they could read these patterns directly from the model's thinking process — even on new patient descriptions the model had never seen before.

Scientists built QArray+, a physics simulator that runs on graphics processors and can model how quantum dot arrays behave in real time—not just at equilibrium. The simulator is 1,000× faster than earlier tools and captures the non-equilibrium dynamics that matter when measuring quantum states quickly, making it practical to generate the training data needed to automate device tuning.

Scientists and engineers often need to calculate how the solutions to differential equations change when they tweak the underlying parameters — a task called computing gradients. GRADSOLVE, a new GPU-accelerated tool, does this 5.6 to 14.1 times faster than existing methods by recording the steps a solver takes and replaying them at fixed intervals, rather than trying to differentiate an adaptive algorithm on the fly.