AI Paper Atlas

Research map / Reinforcement learning / Policy gradient and actor-critic

Goal Reaching with Eikonal-Constrained Hierarchical Quasimetric Reinforcement Learning

Vittorio Giammarino, Ahmed H Qureshi

ICLR 2026 · Reinforcement learning › Policy gradient and actor-critic · approach: Temporal-difference and discounting theory · 1 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
add hierarchy with Eikonal high-level value (Eik-HiQRL over Eik-QRL)Eik-QRL · antmaze navigate giantsuccess rate (%): 8 ± 2 -> 59 ± 7Improved
add hierarchy with Eikonal high-level value (Eik-HiQRL over Eik-QRL)Eik-QRL · antmaze stitch largesuccess rate (%): 23 ± 4 -> 81 ± 8Improved
add Eikonal PDE constraint (Eik-QRL over QRL)QRL · antmaze navigate largesuccess rate (%): 54 ± 9 -> 74 ± 3Improved
use Eikonal constraint instead of HJB constraintHJB-QRL · antmaze navigate mediumsuccess rate (%): 31 ± 4 -> 84 ± 3Improved
add PDE constraints to QRLQRL · pointmaze giant navigatecollision rate (%): 60 ± 8 -> 18 ± 12Improved
add PDE constraints to QRLQRL · pointmazesuccess rate (%)No clear effect
add Eikonal-QRL to HIQL high level (Eik-HiQRL over Eik-HIQL)Eik-HIQL · antsoccer stitch arenasuccess rate (%): 2 ± 0 -> 32 ± 4Improved
Eik-HiQRL versus HIQL on stitching soccerHIQL · antsoccer stitch arenasuccess rate (%): 17 ± 3 -> 32 ± 4Improved
Eik-HiQRL versus HIQL on cube manipulationHIQL · cube-single-playsuccess rate (%): 31 ± 4 -> 12 ± 3Hurt
Eik-HiQRL versus HIQL on scene manipulationHIQL · scene-playsuccess rate (%): 52 ± 3 -> 55 ± 14No clear effect
Eik-HiQRL versus QRL on manipulationQRL · scene-playsuccess rate (%): 8 ± 2 -> 55 ± 14Improved
Eik-HiQRL versus best baseline on humanoidmaze large-stitchbest baseline · humanoidmaze large-stitchsuccess rate: t = 11.7, p ≈ 10^-9Improved

Earlier methods this paper compares with

Named baselines from earlier accepted papers in this paper's results tables: better than 1 of 1 on most shared results.

Matched by method name in results tables read by AI. Authors' own tables tend to favour the newer method, and a comparison can be missed when a method is renamed.

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