Pith Number

pith:SORHSKQO

pith:2026:SORHSKQOXTUTC5ZOJA5WWZRVBY

not attested not anchored not stored refs resolved

Bad Seeing or Bad Thinking? Rewarding Perception for Vision-Language Reasoning

Changpeng Wang, Chong Peng, Fangzhen Lin, Haozhe Wang, Qixin Xu, Taofeng Xue, Wenhu Chen

Vision-language models improve both perception and reasoning by routing rewards to the specific source of error via blindfolded verification.

arxiv:2605.14054 v1 · 2026-05-13 · cs.AI · 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{SORHSKQOXTUTC5ZOJA5WWZRVBY}

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

the root cause of this trade-off is an ambiguity in modality credit assignment: when a VLM fails, is it due to flawed perception (bad seeing) or flawed logic (bad thinking)? ... These techniques are integrated into a Modality-Aware Credit Assignment (MoCA) mechanism, which routes rewards to the specific source of error.

C2weakest assumption

That the blindfolded reasoning proxy in Perception Verification can reliably measure and reward perceptual fidelity independently of reasoning outcomes without introducing new biases or requiring perfect separation of modalities.

C3one line summary

A new RL method called MoCA with Perception Verification rewards perceptual fidelity independently to improve both seeing and thinking in VLMs.

References

100 extracted · 100 resolved · 7 Pith anchors

[1] FirstName LastName , title =

[2] FirstName Alpher , title =

[3] Journal of Foo , volume = 13, number = 1, pages =

[4] Journal of Foo , volume = 14, number = 1, pages =

[5] FirstName Alpher and FirstName Gamow , title =

Receipt and verification

First computed	2026-05-17T23:39:12.607850Z
Builder	pith-number-builder-2026-05-17-v1
Signature	Pith Ed25519 (`pith-v1-2026-05`) · public key
Schema	pith-number/v1.0

Canonical hash

93a2792a0ebce931772e483b6b66350e3792d34ff7ac67fa33c097b01c74cb12

Aliases

arxiv: 2605.14054 · arxiv_version: 2605.14054v1 · doi: 10.48550/arxiv.2605.14054 · pith_short_12: SORHSKQOXTUT · pith_short_16: SORHSKQOXTUTC5ZO · pith_short_8: SORHSKQO

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/SORHSKQOXTUTC5ZOJA5WWZRVBY \
  | 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: 93a2792a0ebce931772e483b6b66350e3792d34ff7ac67fa33c097b01c74cb12

Canonical record JSON

{
  "metadata": {
    "abstract_canon_sha256": "694771751536fcba700787642bf133f0d086ef157772f5033bc5d447655fcf45",
    "cross_cats_sorted": [
      "cs.CV"
    ],
    "license": "http://creativecommons.org/licenses/by/4.0/",
    "primary_cat": "cs.AI",
    "submitted_at": "2026-05-13T19:23:53Z",
    "title_canon_sha256": "9ebacb23d313ac608f8a5c552c8dd22b719d0dee191d005b9d551ca8153f4c7b"
  },
  "schema_version": "1.0",
  "source": {
    "id": "2605.14054",
    "kind": "arxiv",
    "version": 1
  }
}