The AI Performance Plateau: Why Benchmarks Are Losing Their Edge
A new study reveals that the rapid gains seen in artificial intelligence are increasingly limited by benchmark saturation, demanding a rethink of how we measure progress.
Science
Beyond Neural Networks: A New Logic for Graph Data
Researchers are exploring symbolic reasoning as a powerful alternative to message passing in graph neural networks, offering improved expressiveness and interpretability.
Can Code Comments Hide Security Flaws from AI?
New research reveals how subtly crafted comments can sometimes mislead AI-powered code review tools, but also highlights a powerful solution.
Can AI Close the Deal? A New Benchmark for Sales Research
A novel evaluation framework assesses the ability of artificial intelligence to perform effective sales research, revealing significant performance variations between leading models.
Unlocking Insights from Connected Data: A New Theory for Graph-Based Learning
![The study demonstrates that mean squared error, assessed across twenty trials, decreases consistently with increasing [latex]\log(1/\pi)[/latex] for various graph sizes, with performance distinctions observed between graph convolutional networks-both with and without skip connections-and a multilayer perceptron baseline.](https://arxiv.org/html/2602.17115v1/final_figure2.png)
A new theoretical framework clarifies how semi-supervised learning on graphs can effectively leverage network structure with limited labeled data.
Spotting the Fake: New AI Targets Deepfake Videos
Researchers have developed a novel architecture and large-scale dataset to more reliably detect increasingly sophisticated AI-generated video content.
Beyond Actions: Gauging true Financial Guidance with Conversational Benchmarks
A new dataset and evaluation framework, Conv-FinRe, assesses financial recommendation systems by examining alignment with a user’s long-term goals, rather than simply mirroring their past behavior.
Can AI Handle Your Finances? A New Test for Number Skills
Researchers have created a challenging benchmark to assess how well artificial intelligence can reason with numbers in real-world banking tasks.
Can AI Crack Graph Theory’s Toughest Problems?
A new reinforcement learning framework is proving adept at challenging long-held beliefs in extremal graph theory and constructing counterexamples to established results.
Beyond Mean-Variance: AI Takes the Reins of Portfolio Management
![Modern portfolio optimization, traditionally reliant on mean-variance optimization [latex]MVO[/latex], is shown to be outperformed by a reinforcement learning approach [latex]DRL[/latex] during backtesting, suggesting that algorithms mirroring human adaptability-even with inherent imperfections-can navigate market fluctuations more effectively than static, mathematically idealized models.](https://arxiv.org/html/2602.17098v1/figures/backtest.png)
A new study reveals that deep reinforcement learning consistently delivers superior portfolio performance compared to traditional optimization techniques.