arXiv cs.CL· Xinchen Xiao·· 10 天前AI 评分30
Route-Verify-Vote：面向混合域推理的过程条件自一致性框架
Route-Verify-Vote: Procedure-Conditioned Self-Consistency for Mixed-Domain Reasoning
AI 导读
针对 SCoRE 2026 三项训练中未出现的混合域任务，研究者提出 Route-Verify-Vote（RVV）框架，无需更新参数即可让语言模型按域标签选择推理过程、逐项验证并聚合完整答案集。官方测试集上，每题采样 16 个答案集投票的精确集合准确率为 74.6%，自适应 RVV 达 77.3%，在选定域路由上组合多模型后升至 79.4%，最终系统在参赛系统中排名第二。
正文 · 原文
Abstract:Compositional generalization remains challenging when language models must combine familiar reasoning operations in unfamiliar ways. The Scenario-Based Commonsense Reasoning Evaluation (SCoRE) 2026 tests this ability on three mixed domains absent from training and requires models to identify the complete set of correct options for each question.
We introduce Route-Verify-Vote (RVV), a framework for procedure-conditioned self-consistency that uses language models without parameter updates. Route uses the provided domain label to select a reasoning procedure that guides the model in representing and applying the relevant constraints. Verify prompts the model to assess each option against those constraints. Vote aggregates complete answer sets and allocates additional samples to questions with a small vote-count margin between the two most frequent sets. Samples for each question follow the same domain-specific procedure.
On the official test set, voting over 16 sampled answer sets per question achieves an exact-set accuracy of 74.6%. Adaptive RVV reaches 77.3%, and combining models on selected domain routes raises accuracy to 79.4%. The final system ranked second among participating systems. These results support domain-specific reasoning procedures and answer-set disagreement as useful tools for allocating inference-time computation in mixed-domain reasoning.
| Comments: | 12 pages, 8 figures, and 8 tables. Accepted for oral presentation at CCL26-Eval |
| Subjects: | Artificial Intelligence (cs.AI); Computation and Language (cs.CL); Machine Learning (cs.LG) |
| Cite as: | arXiv:2610.08814 [cs.AI] |
| (or arXiv:2610.08814v1 [cs.AI] for this version) | |
| https://doi.org/10.48550/arXiv.2610.08814 arXiv-issued DOI via DataCite |
Submission history
From: Xinchen Xiao [view email]
[v1]
Wed, 23 Sep 2026 16:14:38 UTC (353 KB)
来源：arXiv cs.CL · arxiv.org