pith. sign in
Pith Number

pith:2YNZJQXK

pith:2026:2YNZJQXKERHE44UGTNN3W6KL3R
not attested not anchored not stored refs pending

Reward-Conditioned Reinforcement Learning

Marek Cygan, Michal Nauman, Pieter Abbeel

Conditioning RL agents on reward parameters during single-objective training enables zero-shot adaptation to new rewards via replay data alone.

arxiv:2603.05066 v3 · 2026-03-05 · cs.LG

Add to your LaTeX paper
\usepackage{pith}
\pithnumber{2YNZJQXKERHE44UGTNN3W6KL3R}

Prints a linked badge after your title and injects PDF metadata. Compiles on arXiv. Learn more · Embed verified badge

Record completeness

1 Bitcoin timestamp
2 Internet Archive
3 Author claim open · sign in to claim
4 Citations open
5 Replications open
Portable graph bundle live · download bundle · merged state
The bundle contains the canonical record plus signed events. A mirror can host it anywhere and recompute the same current state with the deterministic merge algorithm.

Claims

C1strongest claim

RCRL improves sample efficiency under the nominal reward parameterization, enables efficient adaptation to new parameterizations, and supports zero-shot behavioral adjustment at deployment.

C2weakest assumption

That recomputing counterfactual rewards from replay data collected under the nominal policy produces unbiased training signals for other reward parameterizations.

C3one line summary

RCRL conditions RL policies on reward parameters and uses shared replay data to train for multiple objectives under a single nominal reward, improving efficiency and adaptability.

Formal links

2 machine-checked theorem links

Receipt and verification
First computed 2026-05-20T01:06:08.893422Z
Builder pith-number-builder-2026-05-17-v1
Signature Pith Ed25519 (pith-v1-2026-05) · public key
Schema pith-number/v1.0

Canonical hash

d61b94c2ea244e4e72869b5bbb794bdc6a3c7c3d2353b502e82050d6f4510a50

Aliases

arxiv: 2603.05066 · arxiv_version: 2603.05066v3 · doi: 10.48550/arxiv.2603.05066 · pith_short_12: 2YNZJQXKERHE · pith_short_16: 2YNZJQXKERHE44UG · pith_short_8: 2YNZJQXK
Agent API
Verify this Pith Number yourself
curl -sH 'Accept: application/ld+json' https://pith.science/pith/2YNZJQXKERHE44UGTNN3W6KL3R \
  | jq -c '.canonical_record' \
  | python3 -c "import sys,json,hashlib; b=json.dumps(json.loads(sys.stdin.read()), sort_keys=True, separators=(',',':'), ensure_ascii=False).encode(); print(hashlib.sha256(b).hexdigest())"
# expect: d61b94c2ea244e4e72869b5bbb794bdc6a3c7c3d2353b502e82050d6f4510a50
Canonical record JSON
{
  "metadata": {
    "abstract_canon_sha256": "c5b4a4bb45cbd5cc2bcec4f144608d501ff9ef6fb5eb15de46ef0744c9be4430",
    "cross_cats_sorted": [],
    "license": "http://creativecommons.org/licenses/by/4.0/",
    "primary_cat": "cs.LG",
    "submitted_at": "2026-03-05T11:29:17Z",
    "title_canon_sha256": "70cb4e591b6b3f80a2eb6f4194501f2e5ff9d1017e8ad8e834565b90df3e761d"
  },
  "schema_version": "1.0",
  "source": {
    "id": "2603.05066",
    "kind": "arxiv",
    "version": 3
  }
}