AdapShot: Efficient Adaptive Many-Shot In-Context Learning
Discover AdapShot, a novel framework boosting Many-Shot In-Context Learning efficiency with adaptive shot optimization and semantic KV cache reuse.
Fast, High-Quality Plan Generation with Self-Improvement AI
Discover how self-improving AI models generate fast, high-quality plans with up to 30% shorter lengths and scalable performance across multiple domains.
Calibrated Moral Reasoning Control in Large Language Models
Explore how calibrated control enhances moral reasoning in large language models using novel techniques for ethical AI decision-making.
Workspace-Bench 1.0: AI Benchmark for Complex File Tasks
Discover Workspace-Bench 1.0, a benchmark for evaluating AI agents on complex workspace tasks with large-scale file dependencies and real-world scenarios.
Real-Time Adversarial Testing of Autonomous Driving Systems
Explore a new framework for real-time evaluation of autonomous driving systems under adversarial attacks using real-world data and advanced learning models...