Interpretable experiential learning based on state history and global feedback

arXiv cs.LG / 5/5/2026

📰 NewsIdeas & Deep AnalysisModels & Research

共有:

Key Points

The paper introduces an interpretable experiential learning model that learns behavioral dynamics as a transition graph over state sets.
Each transition in the graph is annotated with utility and evidence counts, aiming to improve interpretability compared with opaque function approximators.
The approach is designed for reinforcement learning in resource-constrained environments, where model efficiency is critical.
Experiments on the OpenAI Gym Atari Breakout benchmark show performance comparable to some existing neural network-based solutions.
The work is presented as a new arXiv submission, providing an early research contribution that others can build on and compare against.

Abstract

A new interpretable experiential learning model based on state history and global feedback is presented. It is capable of learning a behavioral model represented by a transition graph between sets of states, with transitions attributed with utility and evidence count. This model is expected to be suitable for solving reinforcement learning problem in resource-constrained environments. The model was thoroughly evaluated on the OpenAI Gym Atari Breakout benchmark, demonstrating performance comparable to some known neural network-based solutions.