Dokimos is an evaluation framework for LLM applications in Java. It helps you evaluate responses, track quality over time, and catch regressions before they reach production.
Abstract: Existing approaches often operate poorly in noisy or occluded conditions, rely on cloud-based inference subject to latency and privacy issues, and do not take emotional context into ...
Abstract: This paper presents a novel hybrid framework that combines Dynamic Spanning Tree Coverage (D-STC) with Q-learning to address the challenges of mobile robot navigation and coverage in dynamic ...
Some results have been hidden because they may be inaccessible to you
Show inaccessible results