6 accepted papers
Efficient inference of large language models (LLMs) is hindered by an ever-growing key-value (KV) cache, making KV cache compression a critical research direction. Traditional methods selectively evict less important KV cache entries, which leads to information loss and hallucinations. Recently, mer
Reinforcement learning (RL) in non-stationary environments is challenging, as changing dynamics and rewards quickly make past experiences outdated. Traditional experience replay (ER) methods, especially those using TD-error prioritization, struggle to distinguish between changes caused by the agent'…
Augmented Large Language Models (LLMs) enhance standalone LLMs by integrating external data sources through API calls. In interactive applications, efficient scheduling is crucial for maintaining low request completion times, directly impacting user engagement. However, these augmentations introduce…
Deep Reinforcement Learning (DRL) suffers from uncertainties and inaccuracies in the observation signal in real-world applications. Adversarial attack is an effective method for evaluating the robustness of DRL agents. However, existing attack methods targeting individual sampled actions have limite…
Deep reinforcement learning (DRL) has emerged as a promising approach for robotic control, but its real-world deployment remains challenging due to its vulnerability to environmental perturbations. Existing white-box adversarial attack methods, adapted from supervised learning, fail to effectively t
Fleas in nature generally have potent muscles for jumping and crawling abilities to overcome obstacles in complex environments, while it is challenging for flea-sized robots to have powerful actuators due to the size effect. This work presents a novel high-voltage pulsed actuator for a flea-sized ro