8
Knowledge Graphs and Explainable AI as Complementary Resources for Urban Mining
The paper explores how combining explainable AI techniques with domain knowledge graphs can support human auditors in pre-demolition assessments for urban mining.
8
Google will now disclose which ads are made with AI
Google will begin disclosing when advertisements on its search and video platforms have been generated or altered using artificial intelligence.
8
EQT acquires energy and data center developer Copia Power from Carlyle
Investment firm EQT has acquired Copia Power, a developer of renewable energy and data centers, to expand its digital infrastructure portfolio.
8
China’s drug industry pivots to AI-powered candidates to drive next wave of deals
The Chinese pharmaceutical industry is increasingly integrating artificial intelligence into drug discovery processes to drive future development and investment deals.
8
HPC-Enabled Video-based Coastal Wave Parameter Estimation Using V-JEPA and Deep Spatiotemporal Learning
A deep learning framework utilizing V-JEPA is developed to estimate coastal wave parameters from video data using high-performance computing.
8
Mitigating The Effect of Class Imbalance in Data with Hierarchical and Dependable Structure
This study addresses the challenges of class imbalance in hierarchical cybersecurity vulnerability classification tasks.
8
Beyond Coordinate Gauge: An Audited Protocol for Detecting Donor-Specific Functional Fingerprints after Neural Collapse
This paper introduces a protocol to identify donor-specific functional fingerprints in neural networks that have undergone neural collapse.
8
Evaluating Nonuniform Dependability Across Response Conditions: A Conditional Generalizability Framework Illustrated in Automated Essay Scoring
This study presents a conditional generalizability framework to better evaluate the reliability of automated essay scoring systems across different response conditions.
8
BattVAE-GP: Generative Modeling of Long-Horizon Battery Degradation with Uncertainty Quantification
Researchers developed a hybrid physics-probabilistic model to predict lithium-ion battery degradation across various operating conditions.
8
Burst Spiking Neural Networks
This study proposes Burst Spiking Neural Networks to enhance the accuracy and robustness of low-power neuromorphic computing architectures.
8
OmniPMNet: Bridging discrete and gridded PM10 forecasts via omni-query neural processes
OmniPMNet is a new neural network model designed to integrate station-based and gridded data for more accurate particulate matter forecasting.
8
OOD-RL-Bench: A Benchmark Framework for Out-of-Distribution Detection in Reinforcement Learning
A new benchmark framework, OOD-RL-Bench, is introduced to evaluate out-of-distribution detection capabilities in reinforcement learning agents.
8
When Close Enough Is Not Enough: Autoregressive Drift in Quantum Circuit Synthesis
This study investigates the limitations of transformer-based models in maintaining functional equivalence during quantum circuit synthesis.
8
So Many Opinions, So Many LLMs: Comparing Large Language Models to Traditional Machine Learning for Open- Ended Survey Analysis
This study evaluates the effectiveness of various large language models in analyzing open-ended survey responses compared to traditional machine learning techniques.
8
SHARD: cell-keyed residual splitting for alignment-resistant private dense retrieval
The SHARD framework utilizes cell-keyed residual splitting to provide a trade-off between retrieval performance and document geometry protection.
8
Seeing Through Uncertainty: Free-Energy-Inspired Real-Time Adaptation for Robust Visual Navigation
A new navigation method inspired by the Free Energy Principle enables deep neural networks to adapt more robustly to noisy sensory environments.
8
1D-Bench: A Benchmark for Iterative UI Code Generation with Visual Feedback in Real-World
The 1D-Bench benchmark provides a standardized framework for evaluating AI models that translate UI designs into front-end code.
8
Enabling Energy-Efficient Simultaneous Multi-Task Reinforcement Learning through Spiking Neural Networks with Active Dendrites for Bio-inspired Generalist Agents
Researchers developed an energy-efficient multi-task reinforcement learning approach using spiking neural networks with active dendrites for generalist agents.
8
Higher Embedding Dimension Creates a Stronger World Model for a Simple Sorting Task
An investigation into transformer training reveals that increasing embedding dimensions leads to more robust and consistent internal world models for simple algorithmic tasks.
8
Does Topic Sentiment Cause Perceived Ideology? Comparing Human and LLM Annotations in Political News Articles
A study comparing human and LLM annotations finds that topic sentiment influences perceived political ideology in news articles.
8
Prime Fourier Embeddings: A Principled Basis for Modular Arithmetic
Researchers have developed Prime Fourier Embeddings to help neural networks better represent and process the algebraic structure of integers.
8
Practical Judgment, Virtue, and Intuition in the Use of Opaque AI-Enabled Systems
This paper explores the philosophical and practical challenges of relying on opaque AI systems for autonomous decision-making.
8
恒指午间休盘涨1.46%，恒生科技指数涨1.39%
The Hang Seng Index and Hang Seng Tech Index both rose during midday trading, with AI-related concept stocks like MiniMax and Zhipu seeing notable gains.
8
Optimal Adaptive Market Making: A Theoretical Framework for High-Yield Liquidity Provision in Perpetual Futures Markets
This paper proposes a theoretical framework and stochastic optimal control model for adaptive market making and inventory hedging in perpetual futures markets.