pith:5WKWUVII
VIDA: A dataset for Visually Dependent Ambiguity in Multimodal Machine Translation
A dataset of 2,500 translation instances shows that chain-of-thought fine-tuning helps models use visual evidence to resolve ambiguities more consistently.
arxiv:2605.02035 v2 · 2026-05-03 · cs.CL · cs.AI
Add to your LaTeX paper
\usepackage{pith}
\pithnumber{5WKWUVIIBMRCA3TGGWEMV2LHYP}
Prints a linked badge after your title and injects PDF metadata. Compiles on arXiv. Learn more · Embed verified badge
Record completeness
Claims
Experiments with two state-of-the-art Large Vision Language Models under vanilla inference, supervised fine-tuning (SFT), and our chain-of-thought SFT (CoT-SFT) show that while SFT improves overall translation quality, CoT-SFT yields more consistent gains in disambiguation accuracy, especially on out-of-distribution subsets, indicating a stronger generalization for resolving diverse ambiguity types.
The 2,500 instances are accurately annotated such that visual evidence is genuinely required to resolve each ambiguous span, and the LLM-as-a-judge classifier reliably measures correct span-level disambiguation without its own biases or errors.
VIDA provides 2,500 visually-dependent ambiguous MT instances and LLM-judge metrics; chain-of-thought SFT improves disambiguation accuracy over standard SFT, especially out-of-distribution.
Formal links
Receipt and verification
| First computed | 2026-05-27T01:05:55.900371Z |
|---|---|
| Builder | pith-number-builder-2026-05-17-v1 |
| Signature | Pith Ed25519
(pith-v1-2026-05) · public key |
| Schema | pith-number/v1.0 |
Canonical hash
ed956a55080b22206e663588cae967c3efd4ec7a78fc037fad5db02071830c2a
Aliases
· · · · ·Agent API
Verify this Pith Number yourself
curl -sH 'Accept: application/ld+json' https://pith.science/pith/5WKWUVIIBMRCA3TGGWEMV2LHYP \
| 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: ed956a55080b22206e663588cae967c3efd4ec7a78fc037fad5db02071830c2a
Canonical record JSON
{
"metadata": {
"abstract_canon_sha256": "0ff197cb4b122f311fd9103f89506e2288bc74cba2f1115d0bd45e54eb7e8c97",
"cross_cats_sorted": [
"cs.AI"
],
"license": "http://arxiv.org/licenses/nonexclusive-distrib/1.0/",
"primary_cat": "cs.CL",
"submitted_at": "2026-05-03T19:55:06Z",
"title_canon_sha256": "3dffca7dbaa344ceadcfa408a691b5398bf9917b46db05d0c88f2affc4f09894"
},
"schema_version": "1.0",
"source": {
"id": "2605.02035",
"kind": "arxiv",
"version": 2
}
}