Skip to main content
Search

Radboud University Nijmegen

6 related articles

Subscribe to updates
  1. Before diffusion models generalize, they grow spurious attractors absent from training data but repeated in samples

    As the training set grows, diffusion models move from memorizing training samples to generalizing, and between these phases an intermediate kind of attractor appears that matches no training image yet recurs in generated samples. Reading generation as retrieval in a dense associative memory predicts such spurious states once storage capacity is exceeded; on MNIST, FASHION-MNIST, CIFAR-10 and LSUN-CHURCH the spurious fraction rises and then falls at the memorization-to-generalization boundary, with basin volumes between those of memorized and generalized samples.

  2. arXiv

    Local LLM inference is not private: an EM probe beside a Raspberry Pi recovers generated tokens and prompts

    On a Raspberry Pi 5 running LLaMA-3.2-1B-Instruct locally, electromagnetic leakage captured by a passive probe can identify generated tokens and semantically reconstruct user prompts, showing that local execution alone does not keep prompts and outputs confidential.

  3. arXiv

    One bitstream diffusion recipe spans images, speech and proteins: 3.5% word error in joint speech-text, but two fold-level protein figures disagree

    Continuous bitstream diffusion writes images, speech and proteins as binary codes and, with one training and sampling protocol, reaches 3.5% word error rate in joint speech-text generation, while the fold-level agreement of generated protein sequence-structure pairs is reported as 75% in the results section and 69% in the conclusion.

  4. arXiv

    Agents play full Slay the Spire runs: top models match experts in act one, then fall behind

    Agentic large language models handle complex tasks, yet long-horizon strategic reasoning lacks a dissectable benchmark. SpireBench has 20 LLMs play the complete commercial deck-builder Slay the Spire at its highest difficulty: frontier models beat heuristic baselines and match the average human player but remain far behind experts, and the gap widens as runs lengthen, with the strongest models winning act one as often as experts before falling behind. Memories models write across runs do not improve them and are often wrong and overly cautious.

  5. arXiv

    Treating EEG masking geometry as the only variable: 58 pre-trained models point to a moderate spatial radius with short temporal blocks, and expose a JEPA-specific bias-inflation collapse

    The study unifies EEG self-supervised masking strategies into a three-parameter framework of spatial radius, temporal length and mask ratio, trains 58 models under a fixed REVE-Small backbone and corpus across MAE and JEPA, evaluates them with a linear probe on the 12 OpenEEGBench datasets, and finds that both frameworks agree on a moderate spatial radius with short temporal blocks as the optimum while identifying a JEPA-specific bias-inflation collapse at full-channel masking.

  6. Adaptive Optics Image Analysis Using Generative AI (GPT-4) Scripting: An Exploratory Study

    This exploratory study used GPT-4 to generate and iteratively fine-tune an R script for preprocessing and blob identification/counting in adaptive optics flood illumination ophthalmoscopy (AO-FIO) images, debugging it on images from 4 participants (1 healthy individual and 3 patients with Stargardt disease), having another researcher naive to the prior coding check the code for errors using a different test set of images from 4 other participants (1 healthy individual and 3 patients with Stargardt disease), and comparing cone counts from 5 AO image snippets with counts independently recorded by 2 human graders and those from pre-existing AO analysis software, with the authors positioning the script as functional but nonvalidated.

You’re all caught up