Pith Number

pith:B7QFPS5J

pith:2026:B7QFPS5JZHYBJPHOSH5XHNGHWH

not attested not anchored not stored refs pending

UDM-GRPO: Stable and Efficient Group Relative Policy Optimization for Uniform Discrete Diffusion Models

Chengyuan Wang, Fan Zhang, Haoge Deng, Jiaqi Wang, Ting Pan, Xinlong Wang, Yang Liu, Yonggang Qi

Treating the final clean sample as the action and reconstructing trajectories via the forward process stabilizes reinforcement learning for uniform discrete diffusion models.

arxiv:2604.18518 v3 · 2026-04-20 · cs.CV · cs.LG

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

Add to your LaTeX paper

\usepackage{pith}
\pithnumber{B7QFPS5JZHYBJPHOSH5XHNGHWH}

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

UDM-GRPO significantly improves base model performance across multiple T2I tasks. Notably, GenEval accuracy improves from 69% to 96% and PickScore increases from 20.46 to 23.81, achieving state-of-the-art performance in both continuous and discrete settings. On the OCR benchmark, accuracy rises from 8% to 57%.

C2weakest assumption

The assumption that treating the final clean sample as the action and reconstructing trajectories via the forward process will generalize beyond the specific base models and tasks tested, without introducing new instabilities or overfitting to the chosen benchmarks.

C3one line summary

UDM-GRPO is the first RL integration for uniform discrete diffusion models, using final clean samples as actions and forward-process trajectory reconstruction to raise GenEval accuracy from 69% to 96% and OCR accuracy from 8% to 57%.

Receipt and verification

First computed	2026-05-28T01:04:40.444426Z
Builder	pith-number-builder-2026-05-17-v1
Signature	Pith Ed25519 (`pith-v1-2026-05`) · public key
Schema	pith-number/v1.0

Canonical hash

0fe057cba9c9f014bcee91fb73b4c7b1fbd822e394c870b36a961c071f4b6078

Aliases

arxiv: 2604.18518 · arxiv_version: 2604.18518v3 · doi: 10.48550/arxiv.2604.18518 · pith_short_12: B7QFPS5JZHYB · pith_short_16: B7QFPS5JZHYBJPHO · pith_short_8: B7QFPS5J

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/B7QFPS5JZHYBJPHOSH5XHNGHWH \
  | 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: 0fe057cba9c9f014bcee91fb73b4c7b1fbd822e394c870b36a961c071f4b6078

Canonical record JSON

{
  "metadata": {
    "abstract_canon_sha256": "81eb2ccc25ddb83ad389042ffd99040ef6db22341668918d2c8740503f5f7c74",
    "cross_cats_sorted": [
      "cs.LG"
    ],
    "license": "http://arxiv.org/licenses/nonexclusive-distrib/1.0/",
    "primary_cat": "cs.CV",
    "submitted_at": "2026-04-20T17:16:50Z",
    "title_canon_sha256": "a0c95cb53d0338ac0092cec4c5018c46ea987191444d266a7b3f08d4c37e841e"
  },
  "schema_version": "1.0",
  "source": {
    "id": "2604.18518",
    "kind": "arxiv",
    "version": 3
  }
}