Test-Time Scaling for Small VLMs on Multilingual Visual MCQ

Researchers found that test-time scaling improves small vision-language models on multilingual visual multiple-choice questions, but the largest gains come from fixing prompt parseability and increasing decoding budget rather than from elaborate search or verification methods. Their best configuration achieved 84.1% on the ImageCLEF 2026 test split, ranking first on the Visual MCQ leaderboard.

arXiv:2607.09438v1 Announce Type: new Abstract: Test-time scaling TTS reliably improves reasoning in large language models, but whether it transfers to small open vision-language models remains unclear. We examine this on EXAMS-V, a multilingual visual multiple-choice benchmark, comparing self-consistency, describe-then-reason with PRM-guided beam search, and two post-hoc selectors across Qwen2.5-VL-7B-Instruct and Qwen3.5-4B. What matters is the conditions under which TTS runs, not the search or verification machinery. The largest factor is parseability: an early prompt format left many chains reasoning correctly yet never committing to an answer letter, which a standard answer cue and a guided repair step largely remove. A larger decoding budget removes the rest: raising the per-chain token limit from 1k to 2k recovers 3.7 pp, whereas sampling more chains 8 to 16 adds only 0.15 pp. Once chains have room to finish, elaborate methods contribute little: PRM-guided beam search trails plain self-consistency by 0.39 pp at over eight times the cost, and neither a training-free generative critic nor a trained multimodal PRM beats majority vote across both policies. The largest gain comes instead from the policy model itself +11.4 pp . Our best configuration reaches 84.1% on the held-out ImageCLEF 2026 test split, ranking first on the Visual MCQ leaderboard.