让大模型自动组合推理向量，提升零样本表现与效率
RISER: Orchestrating Latent Reasoning Skills for Adaptive Activation Steering
- 用轻量路由器动态组合可复用的推理向量
- 在7个基准上平均提升3.4%-6.5%零样本准确率
- 适合需要高效可控推理的场景，如智能客服、决策系统
针对大语言模型领域特定推理中参数更新成本高的问题，本文提出RISER（基于路由器的可调推理增强框架），一种无需参数更新的激活空间干预方法。RISER构建可复用的推理向量库，并通过轻量级路由器对每个输入动态组合这些向量。路由器在任务级奖励下通过强化学习优化，以涌现式、组合化方式激活潜在认知单元。在七个多样化基准测试中，RISER相比基线模型平均提升3.4%-6.5%的零样本准确率，且在思维链（CoT）风格推理中实现2-3倍更高的令牌效率，同时获得更稳健的性能增益。进一步分析表明，RISER能自主将多个向量组合为可解释、精确的控制策略，指向更具可控性与高效性的大模型推理路径。
原文摘要 · Abstract (English)
Recent work on domain-specific reasoning with large language models (LLMs) often relies on training-intensive approaches that require parameter updates. While activation steering has emerged as a parameter efficient alternative, existing methods apply static, manual interventions that fail to adapt to the dynamic nature of complex reasoning. To address this limitation, we propose RISER (Router-based Intervention for Steerable Enhancement of Reasoning), a plug-and-play intervention framework that adaptively steers LLM reasoning in activation space. RISER constructs a library of reusable reasoning vectors and employs a lightweight Router to dynamically compose them for each input. The Router is optimized via reinforcement learning under task-level rewards, activating latent cognitive primitives in an emergent and compositional manner. Across seven diverse benchmarks, RISER yields 3.4-6.5% average zero-shot accuracy improvements over the base model while surpassing CoT-style reasoning with 2-3x higher token efficiency and robust accuracy gains. Further analysis shows that RISER autonomously combines multiple vectors into interpretable, precise control strategies, pointing toward more controllable and efficient LLM reasoning.
Thank you to arXiv for use of its open access interoperability. PaperDance 不是 arXiv 官方产品；中文卡片由大模型生成，请以原文为准。