Pith Number

pith:MFIB4ZGC

pith:2025:MFIB4ZGCFJBF3UEPF3T2FITNBN

not attested not anchored not stored refs resolved

Learning to Reason under Off-Policy Guidance

Ganqu Cui, Jianhao Yan, Xiaoye Qu, Yafu Li, Yu Cheng, Yue Zhang, Zhi Wang, Zican Hu

LUFFY mixes off-policy reasoning traces with on-policy rollouts to overcome the limits of standard RLVR in training reasoning models.

arxiv:2504.14945 v5 · 2025-04-21 · cs.LG · cs.AI · cs.CL

Open paper page JSON Open Graph Bundle Merged state Verified badge What is a Pith Number?

Add to your LaTeX paper

\usepackage{pith}
\pithnumber{MFIB4ZGCFJBF3UEPF3T2FITNBN}

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

Compared with previous RLVR methods, LUFFY achieves an over +6.4 average gain across six math benchmarks and an advantage of over +6.2 points in out-of-distribution tasks. Most significantly, we show that LUFFY successfully trains weak models in scenarios where on-policy RLVR completely fails.

C2weakest assumption

That off-policy reasoning traces can be mixed with on-policy rollouts via regularized importance sampling without introducing harmful distribution shift or superficial imitation that would degrade the learned policy.

C3one line summary

LUFFY mixes off-policy reasoning traces into RLVR training via Mixed-Policy GRPO and regularized importance sampling, delivering over 6-point gains on math benchmarks and enabling training of weak models where on-policy RLVR fails.

References

65 extracted · 65 resolved · 18 Pith anchors

[1] OpenAI o1 System Card 2024 · arXiv:2412.16720

[2] DeepSeek-R1: Incentivizing Reasoning Capability in LLMs via Reinforcement Learning 2025 · arXiv:2501.12948

[3] Kimi k1.5: Scaling Reinforcement Learning with LLMs 2025 · arXiv:2501.12599

[4] Chi, Quoc V Le, and Denny Zhou 2022

[5] 7b model and 8k examples: Emerging reasoning with reinforcement learning is both effective and efficient 2025

Formal links

2 machine-checked theorem links

Cited by

28 papers in Pith

Harnessing LLM Agents with Skill Programs

Beyond Mode Collapse: Distribution Matching for Diverse Reasoning

RL-PLUS: Countering Capability Boundary Collapse of LLMs in Reinforcement Learning with Hybrid-policy Optimization

EvoCoT: Overcoming the Exploration Bottleneck in Reinforcement Learning

Sample-efficient LLM Optimization with Reset Replay

Receipt and verification

First computed	2026-05-17T23:38:49.812561Z
Builder	pith-number-builder-2026-05-17-v1
Signature	Pith Ed25519 (`pith-v1-2026-05`) · public key
Schema	pith-number/v1.0

Canonical hash

61501e64c22a425dd08f2ee7a2a26d0b79e76dd1293c56cebf08c0349ba48cd4

Aliases

arxiv: 2504.14945 · arxiv_version: 2504.14945v5 · doi: 10.48550/arxiv.2504.14945 · pith_short_12: MFIB4ZGCFJBF · pith_short_16: MFIB4ZGCFJBF3UEP · pith_short_8: MFIB4ZGC

Agent API

Resolver JSON Graph JSON Events JSON Schema Signing key

Verify this Pith Number yourself

curl -sH 'Accept: application/ld+json' https://pith.science/pith/MFIB4ZGCFJBF3UEPF3T2FITNBN \
  | 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: 61501e64c22a425dd08f2ee7a2a26d0b79e76dd1293c56cebf08c0349ba48cd4

Canonical record JSON

{
  "metadata": {
    "abstract_canon_sha256": "36476bf4f1c0c61dc9393470f17a1c08176e6eed4c404f139a466f0b6b250ab5",
    "cross_cats_sorted": [
      "cs.AI",
      "cs.CL"
    ],
    "license": "http://creativecommons.org/licenses/by/4.0/",
    "primary_cat": "cs.LG",
    "submitted_at": "2025-04-21T08:09:13Z",
    "title_canon_sha256": "a80d72f306e3f138b04b3428df5bee8762a2693c938769e3284fc45c99b69946"
  },
  "schema_version": "1.0",
  "source": {
    "id": "2504.14945",
    "kind": "arxiv",
    "version": 5
  }
}