November 2025 — page 2 of 2
- New Framework Cuts Token Use in AI Retrieval Systems While Boosting Accuracy· Agentic AI
- LLMs can replicate human cooperation patterns in game theory experiments· AI
- Machine Learning Method Reconstructs Flow Across Entire Sewer Networks Using Just 3-4% Sensor Coverage· AI
- New framework predicts depression severity from clinical interviews with uncertainty estimates· AI
- Researchers adapt LLMs to generate music from text, matching human composers in real-world tests· AI
- LLMs Could Screen Recommendation Updates Without Costly Live Tests· AI Research
- Hybrid system combining knowledge graphs, AI, and search agents sets new benchmark for fact-checking accuracy· AI
- New Jailbreaking Technique Exploits Long-Context Models' Safety Blind Spot· AI Security
- Generative AI Reshapes Computational Biology with Better Accuracy, but Faces Data and Scalability Hurdles· AI
- Evolution Strategy Outperforms Gradient Descent for Real-Time Image Generation Control· AI
- LLMs Show Vulnerability to Conspiracy Thinking, Study Finds· AI
- GitHub Copilot Speeds Up Code Tasks But Doesn't Boost Understanding, Study Finds· AI & Education
- Bayesian Framework Measures Uncertainty in LLM Behavior Across Repeated Runs· AI
- New benchmark exposes major gaps in AI's ability to build 3D scenes from text descriptions· AI
- New method trains neural compressors on-the-fly for scientific simulations using sketching· AI
- AI Code Generation Widening Quality Gap—Researchers Propose Model-Based Recovery to Bridge the Gap· AI
- Self-Supervised Learning for Patent Representations Using Section Structure· AI
- New Technique Maps How LLMs Blend Training Knowledge with Context to Generate Explanations· AI
- Re-FORC Predicts Reasoning Quality in Real Time to Cut AI Compute Costs· Agentic AI
- New AI framework helps code-repair systems explain fixes to developers of all skill levels· AI
- New Method Trains Lightweight AI Models to Navigate Screens with Minimal Data· AI
- New Bilingual Benchmark Tests How Well AI Catches Medical Errors in English and Japanese· AI
- Multi-agent AI system uses fine-tuned language models to automate telecom network repairs· AI