AI Paper Atlas

Research map / Reinforcement learning / Visual and representation RL

LISA: Learning Interpretable Skill Abstractions from Language

Divyansh Garg, Skanda Vaidyanath, Kuno Kim, Jiaming Song, Stefano Ermon

NeurIPS 2022 · Reinforcement learning › Visual and representation RL · approach: Skill learning and diffusion policies

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
hierarchical LISA instead of flat Lang DT, 1k demosLang DT · BabyAI GoToSeqsuccess rate (%): 49.3 -> 59.4Improved
LISA instead of flat Lang DT at 100k demosLang DT · BabyAI SynthSeqsuccess rate (%): 64.2 -> 61.2Hurt
LISA instead of flat Lang DT at 100k demosLang DT · BabyAI BossLevelsuccess rate (%): 72.0 -> 69.8Hurt
LISA instead of Lang DT on LOReL with statesLang DT · LOReL Sawyer statessuccess rate (%): 33.3 -> 66.7Improved
LISA instead of Lang DT on LOReL with imagesLang DT · LOReL Sawyer imagessuccess rate (%): 15 -> 40Improved
LISA instead of flat baseline on unseen compositionsFlat · LOReL Sawyer compositionsuccess rate (%): 13.33 -> 20.89Improved
LISA instead of LOReL planner on unseen compositionsLOReL Planner · LOReL Sawyer compositionsuccess rate (%): 18.18 -> 20.89Improved
learned VQ skills instead of k-means skill clusters (ablation)LISA · BabyAI BossLevel 1ksuccess rate (%): 20.2 -> 49.1Improved
skill horizon 1 instead of 5 (ablation)LISA · BabyAI BossLevel 1ksuccess rate (%): 32 -> 52Improved
skill horizon 10 or 50 instead of 5 (ablation)LISA · BabyAI BossLevel 1ksuccess rate (%): 52 -> 47Hurt
increase number of options from 50 to 100 (ablation)LISA · BabyAI BossLevel 1ksuccess rate (%): 47 -> 43Hurt
vary number of options 10 to 50 (ablation)LISA · BabyAI BossLevel 1ksuccess rate (%): 47 -> 47No clear effect

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