alkhder at ImageEval 2026 Shared Tasks: Constrained Selection and Rotation-Averaged Logit Scoring for Multimodal Hallucination Detection

Computer Science > Computation and Language arXiv:2610.22515 (cs) [Submitted on 15 Oct 2026] Title:alkhder at ImageEval 2026 Shared Tasks: Constrained Selection and Rotation-Averaged Logit Scoring for Multimodal Hallucination Detection Authors:Maher Pasa, Maria Hanini, Amro Najjar, Hasan Alkhder

计算机科学 > 计算与语言 arXiv:2610.22515 (cs) [提交于 2026 年 10 月 15 日] 标题：alkhder 参与 ImageEval 2026 共享任务：用于多模态幻觉检测的约束选择与旋转平均 Logit 评分 作者：Maher Pasa, Maria Hanini, Amro Najjar, Hasan Alkhder

Abstract:We describe our submission to Subtask~1b of ImageEval 2026. The task consists of selecting the single visually valid statement from three candidates paired with an image, and is in both English and Modern Standard Arabic (MSA). Our system uses a pre-trained 7B vision—language model in a zero-shot setting, without task-specific training or fine-tuning.

摘要：我们描述了我们提交给 ImageEval 2026 子任务 1b 的方案。该任务包括从与图像配对的三个候选语句中选择唯一视觉上有效的语句，涵盖英语和现代标准阿拉伯语 (MSA)。我们的系统在零样本设置下使用预训练的 7B 视觉-语言模型，无需特定任务的训练或微调。

We formulate the task as a constrained selection problem and choose the answer using next-token log-probabilities. To reduce answer-position bias, we apply a Latin-square rotation to the candidate statements. Our system achieves CI scores of 0.0510 on the English test set and 0.0960 on the MSA test set, reducing CI relative to the official 3B baselines reported on devtest by 83.7% and 80.4%, respectively.

我们将该任务表述为一个约束选择问题，并使用下一个标记的对数概率来选择答案。为了减少答案位置偏差，我们对候选语句应用了拉丁方旋转。我们的系统在英语测试集上实现了 0.0510 的 CI 分数，在 MSA 测试集上实现了 0.0960 的 CI 分数，相较于开发测试集上报告的官方 3B 基线，CI 分别降低了 83.7% 和 80.4%。