Beyond Exact Match: Task-Aware GRPO for Cross-Domain PCBA Visual Question Answering

A multimodal reasoning framework using Task-Aware Group Relative Policy Optimization (GRPO) achieved an Overall Score of 83.24 on the official PCBA Standard-to-Real Grand Challenge leaderboard, according to arXiv paper 2609.21276v1. The framework converts standards-derived, real-world, and auxiliary PCB-domain data into a unified instruction format with verified reasoning traces, and replaces exact-match supervision with multi-component semantic rewards for choice-based questions, distance-aware rewards for counting questions, and an auxiliary format reward. Inference combines answer-option semantic consistency correction, self-consistency voting, and multi-model arbitration to improve robustness in cross-domain Printed Circuit Board Assembly visual question answering.

arXiv:2609.21276v1 Announce Type: new Abstract: In automated Printed Circuit Board Assembly PCBA inspection, standards-guided decisions require systems to jointly reason over fine-grained visual cues, component semantics, and manufacturing knowledge. Although large vision-language models VLMs provide a promising foundation, their deployment is hindered by the domain shift between standards-derived samples and real-world production-line imagery, together with heterogeneous output spaces spanning choice-based and numerical counting tasks. To address these challenges, we propose a multimodal reasoning framework for cross-domain PCBA visual question answering. The framework converts standards-derived, real-world, and auxiliary PCB-domain data into a unified instruction format and constructs verified reasoning traces aligned with visual evidence, question semantics, candidate options, and ground-truth answers. We further introduce Task-Aware Group Relative Policy Optimization GRPO , which moves beyond exact-match supervision by integrating multi-component semantic rewards for choice-based questions, distance-aware rewards for counting questions, and an auxiliary format reward for valid outputs. During inference, answer-option semantic consistency correction, self-consistency voting, and multi-model arbitration are combined to improve prediction robustness. The proposed system achieves an Overall Score of 83.24 on the official PCBA Standard-to-Real Grand Challenge leaderboard, demonstrating the effectiveness of task-aware reward design and robust inference for cross-domain PCBA visual question answering.