Pith Number

pith:AWCZRZQ4

pith:2025:AWCZRZQ4QCF2K4O2MUR2KYFJ4P

not attested not anchored not stored refs resolved

Unified Reward Model for Multimodal Understanding and Generation

Cheng Jin, Hao Li, Jiaqi Wang, Yibin Wang, Yuhang Zang

A single reward model trained jointly on image and video tasks improves preference alignment for both understanding and generation.

arxiv:2503.05236 v2 · 2025-03-07 · cs.CV

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

Add to your LaTeX paper

\usepackage{pith}
\pithnumber{AWCZRZQ4QCF2K4O2MUR2KYFJ4P}

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

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

jointly learning to assess diverse visual tasks yields substantial mutual benefits... achieving consistent improvements across each domain.

C2weakest assumption

The large-scale human preference dataset accurately represents human judgments across tasks and the two-stage filtering strategy produces high-quality, unbiased preference pairs without introducing selection artifacts.

C3one line summary

UnifiedReward is the first unified reward model that jointly assesses multimodal understanding and generation to provide better preference signals for aligning vision models via DPO.

References

64 extracted · 64 resolved · 21 Pith anchors

[1] Diffusion model alignment using direct preference optimization 2024

[2] Videodpo: Omni-preference alignment for video diffusion generation 2024

[4] Lift: Leveraging human feedback for text-to-video model alignment 2024

[5] Llava-critic: Learning to evaluate multimodal models 2024

[6] Internlm-xcomposer2.5-reward: A simple yet effective multi-modal reward model 2025

Formal links

2 machine-checked theorem links

Cited by

26 papers in Pith

Reward Forcing: Efficient Streaming Video Generation with Rewarded Distribution Matching Distillation

Improving Text-to-Image Generation with Intrinsic Self-Confidence Rewards

DSH-Bench: A Difficulty- and Scenario-Aware Benchmark with Hierarchical Subject Taxonomy for Subject-Driven Text-to-Image Generation

Visual-ERM: Reward Modeling for Visual Equivalence

Flow-OPD: On-Policy Distillation for Flow Matching Models

Receipt and verification

First computed	2026-05-18T03:15:18.210251Z
Builder	pith-number-builder-2026-05-17-v1
Signature	Pith Ed25519 (`pith-v1-2026-05`) · public key
Schema	pith-number/v1.0

Canonical hash

058598e61c808ba571da6523a560a9e3c7a30acc1708721a83ab8864b033787e

Aliases

arxiv: 2503.05236 · arxiv_version: 2503.05236v2 · doi: 10.48550/arxiv.2503.05236 · pith_short_12: AWCZRZQ4QCF2 · pith_short_16: AWCZRZQ4QCF2K4O2 · pith_short_8: AWCZRZQ4

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/AWCZRZQ4QCF2K4O2MUR2KYFJ4P \
  | 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: 058598e61c808ba571da6523a560a9e3c7a30acc1708721a83ab8864b033787e

Canonical record JSON

{
  "metadata": {
    "abstract_canon_sha256": "61deb671ab72436b4fe9934b639cef553965b827b86a5f6c7be13ed84a2aa7c3",
    "cross_cats_sorted": [],
    "license": "http://arxiv.org/licenses/nonexclusive-distrib/1.0/",
    "primary_cat": "cs.CV",
    "submitted_at": "2025-03-07T08:36:05Z",
    "title_canon_sha256": "2aedac95cbbedd1322221f81bf6fbba2b5fd14fda6c12942a05f73b48664f4f8"
  },
  "schema_version": "1.0",
  "source": {
    "id": "2503.05236",
    "kind": "arxiv",
    "version": 2
  }
}