Classical Theory Reproduces Radiation Relations, Not Blackbody Spectrum
Preprint: A theoretical analysis finds that classical fields recover links between radiation energy, pressure and flux, but not the spectrum’s absolute normalization.
Search
2,230 results
1521–1540
Flight-control review finds trade-offs among NDI architectures
Preprint: A review compares model-based, sensory, incremental and hybrid nonlinear dynamic inversion, including a Citation II flight-test example.
Mixed training scores higher than next-chunk reasoning in AI tests
An arXiv preprint reports higher post-RLVR scores and sharply lower compute costs for the mixed training route.
Laser study maps how ionized D2O forms a D-D bond
Preprint: Measurements and simulations follow D2+ formation through an indirect pathway with direct, roaming and delayed branches.
DF-MoE tops MAVOS-DD tests, but wider scores vary
An arXiv preprint reports the highest displayed scores across four MAVOS-DD scenarios, while AVLips, cross-dataset and video-only comparisons are less uniform.
New statistical framework adds an inconclusive option to research
A statistical preprint proposes a way to classify estimates into rival possibilities while controlling the risk of assigning the wrong class.
ECG adaptation method reports gains across three test settings
Preprint: The method was tested on three ECG datasets under offline, continual online and independent online adaptation protocols.
Arabic NLP research surges, but dialect coverage remains uneven
An arXiv version 2 preprint dated 25 August 2026 maps 7,120 papers and finds core-task coverage centered on Modern Standard Arabic.
Preprint reports higher benchmark scores for a clinical AI agent
MediSkill-Evo led several fixed benchmark comparisons, but the results do not establish clinical safety or prospective validity.
Small AI Models Transfer Across Domains Without Local Training Text
This preprint reports lower judged error rates for compact models, but its benchmark and language-model evaluation limit the result.
Multimodal training controller leads on two benchmarks but trails on VGGSound
An arXiv version 1 preprint reports strong results on CREMA-D and AV-MNIST, mixed results elsewhere, and no win over VGGSound’s strongest baselines.
A preprint claims a narrow ban on three consecutive powerful numbers
The arXiv preprint targets x^n−1, x^n and x^n+1 when the exponent’s prime factors all fall in a tightly defined class.
Simulated Tasks Reveal a 4.8-Fold Gap in Agent Recovery Costs
A preprint reports that, in the Habitat-Sim simulation, Reflexion and CoPAL had similar success rates but sharply different recovery costs.
Reinforcement learning finds mine-pump designs with fewer simulations
Preprint: In one stochastic mine-pump model, the method reached the highest-reward design after 130 simulations, while Bayesian optimisation outperformed a genetic algorithm on reward summaries.
In one IV simulation, empirical Bayes had the lowest GMM error
Preprint: The method had the lowest error in a baseline simulation and was also applied to schooling data.
AI tutoring answers win more benchmark comparisons after refinement
Preprint: An agent-led process was associated with higher preference rates on held-out Chinese K–12 tutoring tasks, with human checks favoring later answers.
Math preprint presents BBL inequality with separate outputs
Preprint states the bound for −∞ ≤ p ≤ 1/d, with least multiplicative constant C = 1; for m ≥ 2, its power is optimal when C remains 1.
PARTAB reports benchmark gains with less table context
Preprint: PARTAB uses four stages to map a table question to an answer.
AI model rebuilds resting-state brain scans across datasets
Preprint: MnemoDyn showed strong reconstruction and prediction results in tests, but the evaluation did not establish clinical usefulness.
Diffusion models may overfit long before they memorize data
A mathematical preprint reports a best early stopping scale, followed by covariance loss and, at far longer times, collapse of the reverse process onto the training set.