This paper describes LogSigma, our system for SemEval-2026 Task 3: Dimensional Aspect-Based Sentiment Analysis (DimABSA). Unlike traditional Aspect-Based Sentiment Analysis (ABSA), which predicts discrete sentiment labels, DimABSA requires predicting continuous Valence and Arousal (VA) scores on a 1-9 scale. A central challenge is that Valence and Arousal differ in prediction difficulty across languages and domains. We address this using learned homoscedastic uncertainty, where the model learns task-specific log-variance parameters to automatically balance each regression objective during training. Combined with language-specific encoders and multi-seed ensembling, LogSigma achieves 1st place on five datasets across both tracks. The learned variance weights vary substantially across languages due to differing Valence-Arousal difficulty profiles-from 0.66x for German to 2.18x for English-demonstrating that optimal task balancing is language-dependent and cannot be determined a priori.

翻译：本文描述了 LogSigma 系统，该系统用于 SemEval-2026 任务三：维度层面情感分析（DimABSA）。与预测离散情感标签的传统层面情感分析（ABSA）不同，DimABSA 要求在 1-9 的评分范围内预测连续的情感和唤醒度（VA）分数。一个核心挑战在于，情感和唤醒度在不同语言和领域中的预测难度存在差异。我们通过引入学习到的同方差不确定性来应对这一挑战，让模型学习任务特定的对数方差参数，从而在训练过程中自动平衡每个回归目标。结合语言特定的编码器和多随机种子集成方法，LogSigma 在两个子赛道中均获得五个数据集上的第一名。由于语言间情感-唤醒度难度曲线不同（从德语的 0.66 倍到英语的 2.18 倍），学习到的方差权重在不同语言间存在显著差异，这表明最优任务平衡具有语言依赖性，无法事先确定。