AI Paper Atlas

Research map / Reinforcement learning / Visual and representation RL

Factorized Scheduling Principle: Learning Interpretable and Transferable Policies via Structured Additive Functions

Hong Je-Gal, Hyun-Suk Lee

ICML 2026 · Reinforcement learning › Visual and representation RL · approach: In-context and adaptive RL

Read the paperCheck your idea against 137,217 papers

AI-extracted. The measured effects below were read by AI from the paper's full text (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.

Measured effects reported in the paper

What was changedWhereMeasuredEffect
transfer FSP trained at N=10 zero-shot to N=20 instead of using Whittle indexFSP · Wireless schedulingaverage reward: -0.888 -> -0.781Improved
use FSP instead of Whittle index at small NFSP · Wireless scheduling, N=5average reward: 0.096 -> 0.085Hurt
use FSP instead of Whittle indexFSP · Warehouse scheduling, N=20average reward: 0.605 -> 0.649Improved
use FSP instead of Whittle indexFSP · Inventory scheduling, N=20average reward: -0.921 -> -0.726Improved
use FSP instead of Whittle index at training sizeFSP · Inventory scheduling, N=10 (ID)average reward: -0.344 -> -0.353Hurt
use zero-shot FSP instead of per-size trained DQNFSP · Warehouse scheduling, N=10 (ID)average reward: 0.883 -> 0.823Mixed
use FSP instead of per-size trained NAMFSP · Warehouse scheduling, N=5average reward: -9.272 -> 0.630Improved
use per-size trained QR-DQN instead of zero-shot FSPQR-DQN · Inventory scheduling, N=5average reward: -0.206 -> -0.127Mixed
increase candidate count from 4 to 32 for FSP trained at N=8FSP · Synthetic transferranking consistency: 0.9940 -> 0.9941No clear effect
increase candidate count from 4 to 32 for FSP trained at N=8FSP · Synthetic transferreward gap to oracle: 7.9x10^-5 -> 6.15x10^-4Mixed

Closest papers by meaning

Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.

Open scime draft

Get the next scime Dispatch and a monthly field update

Free. We send a confirmation link first; every email has a one-click unsubscribe. Privacy