Can We Trust What AI Tells Us? Gauging the Reliability of Language Models
New research focuses on how well large language models understand their own limitations when generating factual information, particularly biographical details.
Science
Beyond Traditional Methods: Adapting Foundation Models for Smarter Anomaly Detection
A new study explores how pre-trained time series models, combined with efficient fine-tuning techniques, are dramatically improving the accuracy and efficiency of anomaly detection.
Beyond Black Boxes: Ranking Interpretable Machine Learning Methods
A new study offers a rigorous comparison of leading interpretable machine learning techniques, revealing how performance varies across different types of data.
Beyond the ‘Aha!’ Moment: When AI Truly Self-Corrects
New research challenges the notion that large language models experience genuine insight during reasoning, finding that self-correction is rare and only reliably improves performance under conditions of high uncertainty.
The Collective Investment Catch-22
New research reveals that shared investment algorithms inevitably force a choice between maximizing profits and ensuring fair access for all participants.
Predicting the Future of Bond Yields with Machine Learning
New research explores how machine learning models, fueled by macroeconomic data, can accurately forecast ultimate forward rates and improve the precision of bond yield predictions.
When AI Teams Cheat: Lessons from Human Collusion
![Human strategies for preventing collusion are being mapped onto the design of multi-agent artificial intelligence systems, with the goal of creating agents that exhibit similar cooperative and competitive behaviors as humans when faced with shared tasks and limited resources-a process formalized by principles analogous to game theory, where agents maximize their individual utility [latex] U_i [/latex] while simultaneously minimizing the potential for detrimental alliances among competitors.](https://arxiv.org/html/2601.00360v1/mapping_visualization.png)
New research explores how strategies designed to prevent price-fixing and other forms of collusion among humans can be adapted to govern the behavior of multi-agent AI systems.
Beyond Prediction: How Uncertainty Can Boost Investment Returns
A new study reveals that factoring asset-specific prediction uncertainty into portfolio construction consistently improves risk-adjusted performance.
Cooperative Markets: How AI Can Incentivize Liquidity
New research demonstrates a decentralized AI framework that encourages self-interested agents to collectively maximize market liquidity, even with limited communication.
Simple Beats Sophisticated: LSTMs Still Rule Stock Forecasting
New research challenges the prevailing trend toward transformer-based models, demonstrating that standard LSTMs consistently deliver superior stock price predictions.