3 papers
cs.LG2024
Off-Policy Correction For Multi-Agent Reinforcement Learning
MichaÅ Zawalski, BÅażej OsiÅski, Henryk Michalewski +1
Multi-agent reinforcement learning (MARL) provides a framework for problems involving multiple interacting agents. Despite apparent similarity to the single-agent case, multi-agent…
cs.LG2024
Catalytic Role Of Noise And Necessity Of Inductive Biases In The Emergence Of Compositional Communication
Åukasz KuciÅski, Tomasz Korbak, PaweÅ KoÅodziej +1
Communication is compositional if complex signals can be represented as a combination of simpler subparts. In this paper, we theoretically show that inductive biases on both the tr…
cs.AI2024
Subgoal Search For Complex Reasoning Tasks
Konrad Czechowski, Tomasz Odrzygóźdź, Marek ZbysiÅski +5
Humans excel in solving complex reasoning tasks through a mental process of moving from one idea to a related one. Inspired by this, we propose Subgoal Search (kSubS) method. Its k…