AI Paper Atlas

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

SVL: Goal-Conditioned Reinforcement Learning as Survival Learning

Franki NGUIMATSIA TIOFACK, Fabian Schramm, Théotime Le Hellard, Justin Carpentier

ICML 2026 · Reinforcement learning › Policy gradient and actor-critic · approach: Temporal-difference and discounting theory · 0 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
replace TD-learned value (HIQL) with survival value learning (HSVL)hierarchical GCRL · antmaze-giant-navigatesuccess rate (%): 65 -> 74Improved
replace TD-learned value (HIQL) with survival value learning (HSVL)hierarchical GCRL · humanoidmaze-giant-navigatesuccess rate (%): 12 -> 81Improved
replace TD value with survival value on visual taskshierarchical GCRL · visual-antmaze-giant-navigatesuccess rate (%): 6 -> 61Improved
replace TD value with survival value on stitch taskshierarchical GCRL · antmaze-large-stitchsuccess rate (%): 67 -> 31Hurt
replace TD value with survival value on stitch taskshierarchical GCRL · antmaze-medium-stitchsuccess rate (%): 94 -> 83Hurt
use HSVL instead of GCIQL on manipulation puzzleGCIQL · puzzle-3x3-playsuccess rate (%): 95 -> 54Hurt
use survival critic (SVL) instead of contrastive critic (CRL re-eval), flat actorflat DDPG+BC actor · humanoidmaze-giant-navigatesuccess rate (%): 7 -> 45Improved
use survival critic (SVL) instead of contrastive critic (CRL re-eval), flat actorflat DDPG+BC actor · humanoidmaze-large-navigatesuccess rate (%): 42 -> 66Improved
use survival critic (SVL) instead of CRL re-eval, flat actorflat DDPG+BC actor · antmaze-giant-navigatesuccess rate (%): 50 -> 43Hurt
add hierarchical extraction on top of SVL (ablation)SVL · humanoidmaze-giant-navigatesuccess rate (%): 45 -> 81Improved
add hierarchical extraction on top of SVL (ablation)SVL · humanoidmaze-medium/large-navigatesuccess rate (%): 66 -> 66 (large)No clear effect
use PCS hazard instead of finite-horizon or PCH (ablation)HSVL · antmaze/humanoidmaze navigatesuccess rate (%): 71 / 71 / 74 on antmaze-giantNo clear effect

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