AI Paper Atlas

Research map / Reinforcement learning / Visual and representation RL

Hierarchical and Interpretable Skill Acquisition in Multi-task Reinforcement Learning

Tianmin Shu, Caiming Xiong, Richard Socher

ICLR 2018 · Reinforcement learning › Visual and representation RL · approach: Curriculum and environment design · 55 citations (OpenAlex)

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
Use the hierarchical policy instead of a flat policyMinecraft policies trained in small rooms and tested in big rooms · Minecraft big room, Find xsuccess rate: 0.515 -> 0.723Improved
Use the hierarchical policy instead of a flat policyMinecraft policies trained in small rooms and tested in big rooms · Minecraft big room, Get xsuccess rate: 0.450 -> 0.648Improved
Train a hierarchical rather than flat policyGet x policy trained with one item and tested with distracting items · Get x with multiple distracting itemssuccess rate: 29% -> 94%Improved
Add the two-phase curriculum that first trains reuse of base skills (ablation)Hierarchical policy · Minecraft task learningconvergence speedImproved
Add the stochastic temporal grammar (ablation)Hierarchical policy · Minecraft task learninglearning efficiencyImproved
Add alternating updates of the sub-policies (ablation)Hierarchical policy · Minecraft task learninglearning efficiencyImproved
Use a separate switch value function V_k^sw instead of V_k (ablation)Hierarchical policy · Minecraft task learninglearning efficiencyImproved
Use the hierarchical model rather than the fine-tuned flat policyPolicy π1 trained from the same base policy π0 · G1 learningaverage reward: below 0.8Improved
Use the full hierarchical model rather than a flat policyPolicy π1 trained from the same base policy π0 · G1 learningepisodes to convergence: 22,000 episodesImproved

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