Alibaba's HDPO framework trains AI agents to skip unnecessary tool calls, cutting redundant invocations from 98% to 2% while ...

The technique, called Reinforcement Learning with Verifiable Rewards with Self-Distillation (RLSD), combines the reliable ...

Researchers at EPFL have developed 'Synthegy', a framework that uses large language models to evaluate and guide chemical synthesis planning and reaction mechanism analysis through natural-language ...

The ability to make adaptive decisions in uncertain environments is a fundamental characteristic of biological intelligence. Historically, computational ...

Researchers have proposed an integrated eco-driving framework for fuel cell hybrid electric vehicles in multi-lane highway scenarios, using deep reinforcement learning to optimize motion trajectory ...

The rapid rise of electric vehicles combined with breakthroughs in autonomous driving technology is reshaping the future of transportation toward greater sustainability. Intelligent electric vehicles, ...

The bipedal wheel-legged robot combines the high energy efficiency of wheeled movement with the terrain adaptability of legged locomotion. However, achieving a smooth transition between these two ...

Abstract: This paper investigates the dependent task scheduling with service caching (DTSSC) in mobile edge computing (MEC) systems, where each task requires a specific service program for execution.

Abstract: Visual reinforcement learning (VRL) aims to learn optimal policies directly from pixel data, which holds significant potential for applications in control systems characterized by data ...

RLP uses a single network (shared parameters) to (1) sample a CoT policy 𝜋 𝜃 ( 𝑐 𝑡 ∣ 𝑥 < 𝑡 ) π θ (c t ∣x <t ) and then (2) score the next token 𝑝 𝜃 ( 𝑥 𝑡 ∣ 𝑥 < 𝑡 , 𝑐 𝑡 ) p θ (x t ∣x ...