Share your thoughts, 1 month free Claude Pro on usSee more
WorkDL logo mark

AVIS: Adaptive Test-Time Scaling for Vision-Language Models

About

Modern Vision-Language Models (VLMs) benefit from chain-of-thought prompting and test-time scaling, but these gains often come with prohibitive inference cost due to large visual contexts and long decoding chains. We view this cost through two coupled axes: Visual Context Scaling (VCS), which controls how much visual evidence is passed to the language model, and Visual Reasoning Scaling (VRS), which controls how much inference-time reasoning search is performed. Existing methods typically optimize one axis at a time, leaving the joint allocation of compute across these axes underexplored. We introduce Adaptive Visual Inference Scaling (AVIS), a lightweight policy that adapts both VCS and VRS per query. AVIS realizes VCS through Key Diversity Visual (KDV) pruning, a training-free $O(N)$ key-based rule for removing redundant visual tokens before prefilling, and realizes VRS through adaptive self-consistency, using a learned difficulty predictor to select the number of reasoning rollouts. AVIS is deployment-friendly and compatible with shared-prefill inference, where all rollouts reuse a single prefilling pass and KV cache. Across diverse image and video reasoning benchmarks, AVIS improves the accuracy--compute trade-off relative to VCS-only and VRS-only baselines, and remains effective on top of RL post-trained VLMs while keeping compute and latency low.

Ahmadreza Jeddi, Minh Ngoc Le, Amirhossein Kazerouni, Hakki Can Karaimer, Hue Nguyen, Iqbal Mohomed, Michael Brudno, Alex Levinshtein, Konstantinos G. Derpanis, Babak Taati, Radek Grzeszczuk• 2026

Related benchmarks

TaskDatasetResultRank
Multimodal EvaluationMME
Score2.47e+3
902
Multimodal EvaluationMMStar--
177
Object Hallucination EvaluationPOPE
POPE Score89.1
22
General VQAMME
VQA Score2.37e+3
19
Mathematical ReasoningMathVision
Score41.1
12
Mathematical ReasoningMathVista
Score71.8
12
Visual ReasoningCVBench
Score76.1
12
Video VQAVideo-TT
Score38.4
7
Video VQAQ-Bench-Video
Score61.3
7
Image VQAMMMU-Pro
Score43.95
7
Showing 10 of 24 rows

Other info

Follow for update