{"paper":{"title":"SVL: Goal-Conditioned Reinforcement Learning as Survival Learning","license":"http://creativecommons.org/licenses/by-nc-sa/4.0/","headline":"The goal-conditioned value function equals a discounted sum of survival probabilities over time.","cross_cats":["cs.AI"],"primary_cat":"cs.LG","authors_text":"Fabian Schramm, Franki Nguimatsia Tiofack, Justin Carpentier, Th\\'eotime Le Hellard","submitted_at":"2026-04-19T17:44:13Z","abstract_excerpt":"Standard approaches to goal-conditioned reinforcement learning (GCRL) that rely on temporal-difference learning can be unstable and sample-inefficient due to bootstrapping. While recent work has explored contrastive and supervised formulations to improve stability, we present a probabilistic alternative, called survival value learning (SVL), that reframes GCRL as a survival learning problem by modeling the time-to-goal from each state as a probability distribution. This structured distributional Monte Carlo perspective yields a closed-form identity that expresses the goal-conditioned value fun"},"claims":{"count":4,"items":[{"kind":"strongest_claim","text":"This structured distributional Monte Carlo perspective yields a closed-form identity that expresses the goal-conditioned value function as a discounted sum of survival probabilities, enabling value estimation via a hazard model trained via maximum likelihood on both event and right-censored trajectories.","source":"verdict.strongest_claim","status":"machine_extracted","claim_id":"C1","attestation":"unclaimed"},{"kind":"weakest_assumption","text":"That modeling time-to-goal as a probability distribution via a hazard function produces stable value estimates that avoid the bootstrapping instability of temporal-difference methods, and that the three practical estimators (finite-horizon truncation and binned infinite-horizon approximations) faithfully capture long-horizon objectives without introducing significant bias.","source":"verdict.weakest_assumption","status":"machine_extracted","claim_id":"C2","attestation":"unclaimed"},{"kind":"one_line_summary","text":"Survival value learning expresses the goal-conditioned value function as a discounted sum of survival probabilities and estimates it with maximum-likelihood hazard models on censored data, matching or exceeding TD baselines on long-horizon offline GCRL tasks.","source":"verdict.one_line_summary","status":"machine_extracted","claim_id":"C3","attestation":"unclaimed"},{"kind":"headline","text":"The goal-conditioned value function equals a discounted sum of survival probabilities over time.","source":"verdict.pith_extraction.headline","status":"machine_extracted","claim_id":"C4","attestation":"unclaimed"}],"snapshot_sha256":"dda31e33a9c284b4b49ad1db146bed5e837c368ecaf7f8737bada6382368ed50"},"source":{"id":"2604.17551","kind":"arxiv","version":2},"verdict":{"id":"369fb219-d8d7-47ec-8b27-61441b58355b","model_set":{"reader":"grok-4.3"},"created_at":"2026-05-10T06:41:08.833872Z","strongest_claim":"This structured distributional Monte Carlo perspective yields a closed-form identity that expresses the goal-conditioned value function as a discounted sum of survival probabilities, enabling value estimation via a hazard model trained via maximum likelihood on both event and right-censored trajectories.","one_line_summary":"Survival value learning expresses the goal-conditioned value function as a discounted sum of survival probabilities and estimates it with maximum-likelihood hazard models on censored data, matching or exceeding TD baselines on long-horizon offline GCRL tasks.","pipeline_version":"pith-pipeline@v0.9.0","weakest_assumption":"That modeling time-to-goal as a probability distribution via a hazard function produces stable value estimates that avoid the bootstrapping instability of temporal-difference methods, and that the three practical estimators (finite-horizon truncation and binned infinite-horizon approximations) faithfully capture long-horizon objectives without introducing significant bias.","pith_extraction_headline":"The goal-conditioned value function equals a discounted sum of survival probabilities over time."},"integrity":{"clean":true,"summary":{"advisory":0,"critical":0,"by_detector":{},"informational":0},"endpoint":"/pith/2604.17551/integrity.json","findings":[],"available":true,"detectors_run":[],"snapshot_sha256":"c28c3603d3b5d939e8dc4c7e95fa8dfce3d595e45f758748cecf8e644a296938"},"references":{"count":0,"sample":[],"resolved_work":0,"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57","internal_anchors":0},"formal_canon":{"evidence_count":0,"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"author_claims":{"count":0,"strong_count":0,"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"builder_version":"pith-number-builder-2026-05-17-v1"}