Science – Page 76

Beyond Whole Numbers: Neural Networks Tackle Fractional Growth

24.11.2025 by qfx

$Fractional exponential growth demonstrates that a quantity can increase at a rate proportional to a non-integer power of itself, offering a nuanced alternative to traditional exponential models where growth is governed by $e^x$ and instead allowing for rates described by functions like $e^{x^\alpha}$ with $0 < \alpha < 1$.$

A novel approach combines artificial neural networks with fractional calculus to more accurately model complex growth patterns.

Decoding Market Noise: Can Financial Data Generate true Randomness?

24.11.2025 by qfx

$The Hamming correlation, computed over sequences of length 32, demonstrates a quantifiable relationship between input patterns, effectively measuring their similarity through bitwise comparison-a fundamental operation in information theory encapsulated by the formula $H(x, y) = \sum_{i=0}^{L-1} (x_i \oplus y_i)$, where $L$ represents the sequence length and $\oplus$ denotes the exclusive OR operation.$

New research explores whether the inherent unpredictability of high-frequency financial markets can be harnessed to create statistically sound random number sequences.

Spotting the Fake: AI Image Detection Gets a Contrastive Boost

24.11.2025 by qfx

Model performance in source attribution varies predictably with the number of few-shot samples provided per generator, demonstrating that accuracy increases as more examples are available to each generator during training.

A new approach using contrastive learning significantly improves the detection of AI-generated images and identifies their origins, even with limited examples.

Smarter Chip Layouts: AI-Powered Multipatterning Takes Shape

24.11.2025 by qfx

The methodology decomposes graph coloring into two sequential phases-an initial solution generation followed by a refinement stage-to effectively address the problem.

A new approach uses graph neural networks to automatically balance and optimize the complex patterns needed for modern microchip design.

Sharper Views of the Universe: AI-Powered Denoising of Weak Lensing Maps

24.11.2025 by qfx

New research demonstrates that advanced artificial intelligence techniques can significantly improve the clarity of weak gravitational lensing maps, paving the way for more precise cosmological measurements.

Outsmarting AI: Defending Against Prompt Hijacking

23.11.2025 by qfx

A multi-layered defense framework systematically scrutinizes retrieved content through embedding analysis, content filtering, and guardrail application, culminating in response verification before any output is generated - a process acknowledging that even robust systems are ultimately vulnerable to unforeseen production edge cases.

New research details a robust framework for protecting AI agents from increasingly sophisticated prompt injection attacks, ensuring reliable and secure operation.

Reviving the Past: AI-Powered Mapmaking for Historical Documents

23.11.2025 by qfx

Synthetic historical maps are generated and then subjected to semantic segmentation, demonstrating a methodology for both recreating and analyzing cartographic data.

A new approach leverages generative AI to automatically create training data for accurately segmenting and understanding centuries-old maps.

Banking on Attention: Personalizing Finance with Transformers

23.11.2025 by qfx

$Sequential user representations, $Su^\hat{S\_{u}}$, are constructed by processing dynamic context ($FdF\_{d}$) and clickstream sequences ($SuS\_{u}$), and then temporally aligned with both static features ($FsF\_{s}$) and field-modulated embeddings ($FfmF\_{fm}$) to create a comprehensive user profile.$

A new framework leverages the power of sequential data modeling to deliver more relevant advertising and services in the financial sector.

Smart Data for Smarter Robots: Distilling Knowledge from Vast Experience

23.11.2025 by qfx

A new framework efficiently compresses large vision-language-action datasets into smaller, more manageable sets for training robot learning models.

Spiking Neurons Tune In: AI for Cleaner Radio Signals

22.11.2025 by qfx

A unified spiking neural network, trained as a large model and subsequently fragmented, facilitates radio frequency interference detection across diverse neuromorphic platforms-including both snnTorch simulations and SynSense Xylo hardware-with latency encoding of spectrograms and deployment in a standardized NIR format enabling comparative power measurements.

Researchers have developed an end-to-end artificial intelligence pipeline leveraging spiking neural networks to effectively filter out unwanted radio interference and enhance the clarity of astronomical data.