Emo-Jev：基于Jev概率推理的情感分类
原标题：Emo-Jev: Probabilistic Reasoning for Emotion Classification with Jev
AI 摘要
该论文提出 Emo-Jev，一个无需训练的框架，利用 Jev 的概率决策接口进行情感、情绪、讽刺和幽默分类。Emo-Jev 包含两种实现：Emo-Jev-D 将分类分解为原子判断并组合概率，Emo-Jev-SC 构建多路径判断并聚合为共识决策。在八个数据集上，标准 Jev 平均 macro-F1 为 62.93%，低于最强 LLM 基线 Gemini 的 67.28%，但推理成本低 96.41%，中位延迟仅 250 毫秒；两种 Emo-Jev 变体得分分别为 62.00% 和 61.61%，未稳定超越直接分类。
正文节选
Emo-Jev: Probabilistic Reasoning for Emotion Classification with Jev Abstract Jev offers an alternative interface for language understanding: given an input and predefined questions, it returns probabilistic decisions rather than free-form responses. Whether this interface can support effective reasoning for text classification against leading LLMs remains an open questions. We introduce Emo-Jev, a training-free framework with two complementary implementations. Emo-Jev-D decomposes classificatio