successor-representation

A reusable framework for successor features for transfer in deep reinforcement learning using keras.

Deep Successor Representation

Code for the paper Optimistic Linear Support and Successor Features as a Basis for Optimal Policy Transfer - ICML 2022

Python implementation of SR-Dyna from Russek, Momennejad et al 2017

Cross-session memory for MCP agents — episodic + semantic, recall ranked by a successor-representation prior.

Implementing the DCEO Algorithm (with some modifications!) from Klissarov and Machado (2023).

TMLR 2026: Constrained Reinforcement Learning Using Successor Representations

Is the shadow of the future cached as value or as prediction? Computational modeling of learning in repeated cooperation (reanalysis of Dal Bo & Frechette, 2011).

[HuMoWM@ECCV2026] Executable Behaviour Representations

Mental maps help people plan, but leading models bias these maps toward past choices. Here, we show that reweighting past experiences lets a model learn reusable maps that capture both the flexibility and biases of human replanning.