Researchers have developed a method to formalize implicit social norms into quantifiable principles, enabling AI agents to better coordinate with humans. In pedestrian-vehicle interaction experiments,…
A new study with 968 participants and 2,904 conversations found that AI language models generate responses perceived as more empathic than those written by humans, but people feel less heard when they…
Researchers introduced ANYPoC, a multi-agent framework that automates bug validation by generating executable proofs of concept. The system outperformed state-of-the-art coding agents like Claude Code…
ContestTrade, a new AI-driven trading system, uses internal competition between AI agents to refine financial decisions, outperforming traditional methods in post-2024 A-share backtests with superior …
Researchers analyzed 2.6 billion sketches from 236 countries to study how visual representations of concepts vary across cultures, finding that visual similarities align 45% more with cultural distanc…
Researchers have developed a multi-agent architecture for large language models that factorizes roles into delegation, execution, and answer generation, finding that scaling the delegation role improv…
Researchers introduced SAMPA, a Whisper-based model for prosodic segmentation in Brazilian Portuguese, achieving F1 scores of 0.731 on held-out test data and 0.796 on the MuPe-Diversidades dataset. Th…
Researchers introduced Latent Policy Steering (LPS), a new offline reinforcement learning method that directly integrates Q-gradients from the original action space into latent-space optimization, ach…
AI agents, operating as digital employees, require security protocols akin to software and management practices similar to human workers, according to a new analysis. Businesses must integrate these a…
The AI industry faces a deepening crisis as demand for skilled analysts surges while supply remains critically low, risking stalled projects and eroded confidence. Without urgent investment in educati…
Researchers have developed WISER, a watermark detection algorithm that identifies and localizes AI-generated text with high speed and accuracy, outperforming existing methods. The tool uses epidemic c…
Scientists developed a label-efficient AI model using Siamese neural networks that predicts phase transitions in physics with minimal data, achieving accuracy on par with established methods without r…
Researchers have developed a new numerical method that combines neural networks with Wiener chaos decomposition and the Euler scheme to solve backward stochastic differential equations (BSDEs), offeri…
Researchers have developed new methods to predict the true risk of deep learning models using disagreement-based certificates, which provide tighter generalization guarantees without modifying the tar…
Researchers have developed a new method for neural network depth adaptation that uses error estimation from continuous-time optimal control problems to dynamically insert layers where errors peak. The…
Researchers introduced the Riemann-normal-coordinate Levenberg-Marquardt (RNC-LM) method for nonlinear optimization, which improves convergence and robustness by incorporating geodesic acceleration to…
Researchers found that the Gemma 4 2B vision-language model achieved 93.65% accuracy in detecting fast radio bursts, nearly matching the specialized SwinYNet detector's 92.90%, while producing far few…
Researchers have developed a spectral preprocessing method using FFT that significantly improves attention mechanisms in character-level language models, achieving up to a 79% reduction in validation …
Researchers have developed a new framework combining reinforcement learning with model predictive control to ensure safety during the learning phase of complex cyber-physical systems. The approach use…
Researchers introduced GeoSD, a geometric self-distillation method for language model training that reduces drift in on-policy distillation. The technique improves out-of-distribution accuracy by 5.7 …