Track papers across labs, topics, and trends.
Search papers, labs, and topics across Lattice.
This paper introduces LOCI, a novel framework that enhances Vision-Language Models (VLMs) by decoupling the processes of visual search and evidence verification through a Locator-Critic architecture. The framework employs a Locator agent to identify potential visual evidence and a Critic agent to assess its relevance, iteratively refining the evidence to improve accuracy. LOCI achieves significant performance improvements, setting new state-of-the-art results across various complex visual benchmarks, including substantial gains for both open-weight and proprietary models.
VLMs can now achieve state-of-the-art accuracy by decoupling visual search from evidence verification, revealing the critical role of perceptual grounding in complex tasks.
Vision-Language Models (VLMs) still struggle on tasks requiring complex visual understanding. We argue that the core issue is not high-level reasoning, but instead failing to locate critical details in the image. Due to this shortcoming, VLMs generate often plausible but incorrect reasoning based on flawed perceptual grounding. To address this, we propose Locator-Critic (LOCI), a training-free framework that decouples visual search from evidence verification. LOCI employs a Locator agent to propose candidate visual evidence and a separate Critic agent to evaluate its relevance and sufficiency. These agents engage in an iterative refinement loop, progressively improving the evidence until it is adequate to answer the given question. This decoupled, self-correcting process yields substantial performance gains, achieving state-of-the-art results on multiple complex visual benchmarks. LOCI improves accuracy for both open-weight models like Qwen3-VL (+12.1 on V*, +5.8 on HR-Bench and +11.2 on VisualProbe-Hard) and proprietary models like Gemini 2.5 Pro (+8.9 on V*, +4.3 on HR-Bench, +4.8 on VisualProbe-Hard).