Large Language Models (LLMs) natively default to literal semantic interpretations, making zero-shot irony detection a persistent challenge. We introduce the Robust Dual-Signal (RDS) Fusion framework, a hybrid neuro-symbolic architecture that compresses Chain-of-Thought (CoT) reasoning trajectories without Supervised Fine-Tuning (SFT). Evaluated on a strictly held-out TweetEval test set (N=734), RDS achieves 78.1% accuracy and a Macro F1 of 0.777, matching the absolute performance ceiling of the fine-tuned BERTweet. On the heavily imbalanced iSarcasm dataset, the frozen CoT pipeline filters 22.5% of out-of-distribution hallucinations, yielding a zero-shot Macro F1 of 0.6726 and Ironic F1 of 0.4821, outperforming multiple heavily supervised SemEval transformer ensembles. A statistical ablation confirms this structural synergy: adding the symbolic prior to the neural baseline yields no significant gain (p = 0.242), and the marginal benefit of adding the CoT pipeline to that prior is heavily compressed (p = 0.149). Only the complete, concurrent fusion of all three signals achieves a statistically validated improvement over the baseline (p = 0.005).

翻译：大语言模型（LLMs）天然默认为字面语义解释，这使得零样本讽刺检测成为一个持久性挑战。我们提出鲁棒双信号（RDS）融合框架，这是一种混合神经-符号架构，可在无需监督微调（SFT）的情况下压缩链式思维（CoT）推理轨迹。在严格保留的TweetEval测试集（N=734）上，RDS实现了78.1%的准确率和0.777的宏F1分数，与经过微调的BERTweet模型的绝对性能上限持平。在高度不平衡的iSarcasm数据集上，冻结的CoT流水线过滤了22.5%的分布外幻觉，实现了0.6726的零样本宏F1分数和0.4821的讽刺类F1分数，优于多个强监督SemEval Transformer集成模型。统计消融实验证实了这种结构协同效应：将符号先验加入神经基线模型未产生显著提升（p = 0.242），而向该先验加入CoT流水线的边际收益被高度压缩（p = 0.149）。只有三个信号完整而并发的融合才能实现对基线的统计显著改进（p = 0.005）。