Pith Number

pith:LFHXT7SD

pith:2026:LFHXT7SDAKAV62UOXQYQ4XBFOF

not attested not anchored not stored refs resolved

Pareto-Guided Optimal Transport for Multi-Reward Alignment

Bing Su, Guiwei Zhang, Ji-Rong Wen, Mohan Zhou, Tianyu Zhang, Wenyi Mo, Yalong Bai, Ying Ba

PG-OT builds prompt-specific Pareto frontiers and applies distribution-aware optimal transport to improve multi-reward alignment while introducing JDR and JCR metrics to measure synergy and hacking.

arxiv:2605.13155 v1 · 2026-05-13 · cs.CV

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

Add to your LaTeX paper

\usepackage{pith}
\pithnumber{LFHXT7SDAKAV62UOXQYQ4XBFOF}

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

Experimental results show that our approach outperforms strong baselines with an 11% gain in JDR and achieves a near 80% win rate in human evaluations.

C2weakest assumption

That a prompt-specific Pareto frontier can be constructed reliably from the available reward models and that mapping samples to it via optimal transport will consistently reduce reward hacking without introducing new instabilities or excessive compute cost.

C3one line summary

PG-OT builds prompt-specific Pareto frontiers and applies distribution-aware optimal transport to improve multi-reward alignment while introducing JDR and JCR metrics to measure synergy and hacking.

References

49 extracted · 49 resolved · 2 Pith anchors

[1] Scaling Learning Algorithms Towards

[2] and Osindero, Simon and Teh, Yee Whye , journal =

[3] Deep learning , author=. 2016 , publisher= 2016

[4] ReNO: Enhancing One-step Text-to-Image Models through Reward-based Noise Optimization , booktitle = 2024

[5] URL https://doi.org/10.1109/CVPR52733 2024 · doi:10.1109/cvpr52733.2024.00763

Formal links

2 machine-checked theorem links

Receipt and verification

First computed	2026-05-18T03:08:57.039355Z
Builder	pith-number-builder-2026-05-17-v1
Signature	Pith Ed25519 (`pith-v1-2026-05`) · public key
Schema	pith-number/v1.0

Canonical hash

594f79fe4302815f6a8ebc310e5c257150d1f4823be94dcbbad9794a4b625074

Aliases

arxiv: 2605.13155 · arxiv_version: 2605.13155v1 · doi: 10.48550/arxiv.2605.13155 · pith_short_12: LFHXT7SDAKAV · pith_short_16: LFHXT7SDAKAV62UO · pith_short_8: LFHXT7SD

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/LFHXT7SDAKAV62UOXQYQ4XBFOF \
  | 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: 594f79fe4302815f6a8ebc310e5c257150d1f4823be94dcbbad9794a4b625074

Canonical record JSON

{
  "metadata": {
    "abstract_canon_sha256": "09a100b4be289c28b6f99800e16e9cc786c7102764f4113fccb3eabc585c2a7d",
    "cross_cats_sorted": [],
    "license": "http://arxiv.org/licenses/nonexclusive-distrib/1.0/",
    "primary_cat": "cs.CV",
    "submitted_at": "2026-05-13T08:19:48Z",
    "title_canon_sha256": "5a5506361d91dc6f897e4f2fb9d72678516c4105ca922e9092d83b66a7b4dd06"
  },
  "schema_version": "1.0",
  "source": {
    "id": "2605.13155",
    "kind": "arxiv",
    "version": 1
  }
}