Yoav Ger

Are world models necessary to achieve human-level agents, or is there a model-free short-cut?
Our new #ICML2025 paper tackles this question from first principles, and finds a surprising answer, agents _are_ world models… 🧵
arxiv.org/abs/2506.01622

04.06.2025 15:48 — 👍 42    🔁 15    💬 2    📌 4

Learning Dynamics of RNNs in Closed-Loop Environments Recurrent neural networks (RNNs) trained on neuroscience-inspired tasks offer powerful models of brain computation. However, typical training paradigms rely on open-loop, supervised settings, whereas ...

All our motor control modelling efforts focus on closed-loop systems for this reason:

"...closed-loop and open-loop training produce fundamentally different learning dynamics, even when using identical architectures and converging to the same final solution."

arxiv.org/abs/2505.13567

29.05.2025 17:28 — 👍 23    🔁 5    💬 0    📌 0

מדויק

31.12.2024 09:54 — 👍 1    🔁 0    💬 0    📌 0