Two language models outperform Wazuh and OpenSearch in a lab test
A preprint study found Meta Llama and Qwen detected more Linux authentication anomalies, while GPT-OSS often failed to return valid structured output.
2041–2060
A preprint study found Meta Llama and Qwen detected more Linux authentication anomalies, while GPT-OSS often failed to return valid structured output.
Preprint: A finite-basis calculation follows Chern-number changes across weak, intermediate and strong periodic-potential regimes.
Preprint: A mathematical model links the relations' stability to fading interior information and finds larger departures around sharp density jumps.
A preprint reports excellent agreement for diffusion and good agreement at short scattering times, but two-qubit tests failed after a few Trotter steps.
Preprint: A high-energy gravity model yields a momentum-kick series with zero convergence radius; its Borel sum is identified with a finite integral when alpha1 is below two.
Preprint: Calculations suggest a light–atom protocol can generate SWAP-alpha gates across arbitrary logical dimensions, but no experiment tested it.
Preprint measurements at two metal interfaces found shorter-lived image-potential-state signals on the more strongly corrugated layer, while the substrates differed in other ways too.
Theoretical work derives exact generating-function identities for plane hypermaps with mixed boundaries, while leaving broader cases open.
Preprint: A mathematical paper builds a connected example with multiple good involutions and shows when these structures split into component-level choices.
Preprint: Formal results show different fairness measures become computationally difficult at different horizons.
A preprint comparing people, macaque brain signals and AI architectures finds that strong scores on ordinary videos can hide weak motion representations.
Preprint: A laboratory test reported lower probe disturbance and nearly two orders of magnitude of coherent technical-noise suppression.
Preprint: An analysis of one 2019 event reports GNSS signal fluctuations, but leaves their cause and navigation impact unresolved.
Preprint: A three-dataset evaluation found weaker external performance for TB cough models, while a clinical-variable baseline transferred more consistently.
An arXiv preprint links feedback from zebrafish body bending with swimming coordination, using live-fish experiments, simulations and a robot.
Preprint replay of stored benchmark pools found judge-guided selection was higher when correct candidates were rare or intermediate, but lower when they were common.
Preprint: A retrospective modeling study reported higher held-out scores than comparison models and similar average performance after clinician-defined restrictions were applied.
An arXiv preprint dated 26 August 2026 tests how much race and age differences could arise under exact fairness before an intervention is chosen.
Preprint: A theoretical framework links differential privacy to exploration, exploitation and firm-wide budget allocation, but its evidence is simulation-based.
Preprint testing a Green-function method finds close agreement with analytical and finite-element calculations in a two-dimensional cylindrical model.