Claims

Response-format reliability is a distinct evaluation concern, with invalid response rates varying independently of model scale

Claim · All claims
Evidence · 1 study · design r2 · q2
Currency · newest study 2026 · not yet checked against its sources · not reviewed by a person · recheck by 2029: tool and model accuracy changes quickly

Subclaims

Reading the evidence codes

q is the design tier, i the impact, n the sample, and the kind with its r the rigour judged within that kind: what each code means.

This claim: no standardized effect is recorded in observations/ for it yet, so no benchmark reading is possible.

q2 i? Invalid response rates across the 30 evaluated VLMs are highly skewed: most models fall below 1%, but Yi-VL-6b and DeepSeek-VL2-Tiny exceed 12%, and even GPT-5.6-Sol retains a 2.62% invalid rate, so response-format reliability constitutes a distinct evaluation concern alongside answer correctness. → Luyao Zhu 2026

Evidence

Luyao Zhu 2026

Luyao Zhu, Xun Wei Yee, Wei Li, Mun Thye Mak, Wee Siong Ng. (2026). MUSE: Benchmarking Large Vision-Language Models on Multi-modal Understanding in Situated Education. arXiv preprint. https://arxiv.org/abs/2609.19088 doi:10.48550/arxiv.2609.19088

q2 · i? · design · r2

Invalid response analysis across the 30 evaluated models (Figure 13), showing a highly skewed distribution. The article reports that Yi-VL-6b and DeepSeek-VL2-Tiny exceed 12% invalid rates while most models are below 1%, and that invalid rates are not determined solely by model scale.

"Invalid responses are not determined solely by model scale: models within the same family vary substantially, and GPT-5.6-Sol retains a 2.62% invalid rate despite its strong task performance."

Discussion

Page metadata
Field Value
type claim
id invalid-response-rates-distinct-evaluation-concern
title Response-format reliability is a distinct evaluation concern, with invalid response rates varying independently of model scale
description Response-format reliability is a distinct evaluation concern, with invalid response rates varying independently of model scale
status review
evidence_strength weak
newest_evidence 2026
recheck_by 2029
generated by process:wiki-ingest on 2026-10-11
trust unverified — no one has confirmed this page's content
sources 1 — luyao-zhu-2026

Frontmatter is the page's machine-readable record — see the Schema & Guide.