Pith Number

pith:H2ZNH2MV

pith:2026:H2ZNH2MVF5EULZMDPIEV6I7DEZ

not attested not anchored not stored refs pending

The Expense of Seeing: Attaining Trustworthy Multimodal Reasoning Within the Monolithic Paradigm

Dikshant Kukreja, Karan Goyal

Vision-language models bypass visual input using language priors, with the penalty increasing as language models scale.

arxiv:2604.20665 v2 · 2026-04-22 · cs.CV · cs.AI

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

Add to your LaTeX paper

\usepackage{pith}
\pithnumber{H2ZNH2MVF5EULZMDPIEV6I7DEZ}

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

state-of-the-art models frequently exhibit functional blindness, i.e., exploiting strong language priors to bypass severe visual representation bottlenecks... hypothesising that as the underlying language engines scale to unprecedented reasoning capabilities, the mathematical penalty of the visual knowledge bottleneck paradoxically increases.

C2weakest assumption

That the Modality Translation Protocol can isolate architectural incapacity from dataset biases without introducing its own translation artifacts or new priors, and that the proposed metrics validly quantify the visual bottleneck.

C3one line summary

Vision-language models exhibit functional blindness by exploiting language priors over visual representations; the Modality Translation Protocol and metrics like Toll, Curse, and Fallacy of Seeing reveal this, supporting a Divergence Law where larger language models increase the visual penalty.

Receipt and verification

First computed	2026-05-22T01:04:02.885254Z
Builder	pith-number-builder-2026-05-17-v1
Signature	Pith Ed25519 (`pith-v1-2026-05`) · public key
Schema	pith-number/v1.0

Canonical hash

3eb2d3e9952f4945e5837a095f23e32677716bfe0b25edca762fe6cf458376fa

Aliases

arxiv: 2604.20665 · arxiv_version: 2604.20665v2 · doi: 10.48550/arxiv.2604.20665 · pith_short_12: H2ZNH2MVF5EU · pith_short_16: H2ZNH2MVF5EULZMD · pith_short_8: H2ZNH2MV

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/H2ZNH2MVF5EULZMDPIEV6I7DEZ \
  | 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: 3eb2d3e9952f4945e5837a095f23e32677716bfe0b25edca762fe6cf458376fa

Canonical record JSON

{
  "metadata": {
    "abstract_canon_sha256": "67b23d5c0fdb88816e23c3f95679819c77e6286bde38d88f32b15801dd2cdd11",
    "cross_cats_sorted": [
      "cs.AI"
    ],
    "license": "http://creativecommons.org/licenses/by-sa/4.0/",
    "primary_cat": "cs.CV",
    "submitted_at": "2026-04-22T15:15:32Z",
    "title_canon_sha256": "f3cf5d5fc844f5b1f4922f230e263fa382523ebddaa753ffad22d1717572d702"
  },
  "schema_version": "1.0",
  "source": {
    "id": "2604.20665",
    "kind": "arxiv",
    "version": 2
  }
}