A survey of temporal credit assignment in deep reinforcement learning

E Pignatelli, J Ferret, M Geist, T Mesnard… - arxiv preprint arxiv …, 2023 - arxiv.org
The Credit Assignment Problem (CAP) refers to the longstanding challenge of
Reinforcement Learning (RL) agents to associate actions with their long-term …

Action Gaps and Advantages in Continuous-Time Distributional Reinforcement Learning

H Wiltzer, MG Bellemare, D Meger, P Shafto… - arxiv preprint arxiv …, 2024 - arxiv.org
When decisions are made at high frequency, traditional reinforcement learning (RL)
methods struggle to accurately estimate action values. In turn, their performance is …

Assessing the Zero-Shot Capabilities of LLMs for Action Evaluation in RL

E Pignatelli, J Ferret, T Rockäschel… - arxiv preprint arxiv …, 2024 - arxiv.org
The temporal credit assignment problem is a central challenge in Reinforcement Learning
(RL), concerned with attributing the appropriate influence to each actions in a trajectory for …

Improving llm generation with inverse and forward alignment: Reward modeling, prompting, fine-tuning, and inference-time optimization

H Sun, T Pouplin, N Astorga, T Liu… - The First Workshop on … - openreview.net
Large Language Models (LLMs) are often characterized as samplers or generators in the
literature, yet maximizing their capabilities in these roles is a complex challenge. Previous …

Credit Assignment in Deep Reinforcement Learning

T Mesnard - 2023 - theses.hal.science
Deep reinforcement learning has been at the heart of many revolutionary results in artificial
intelligence in the last few years. These agents are based on credit assignment techniques …