NeuReasoner: Theory-grounded Mapping of Reasoning Elicitation Boundaries

📄 arXiv: 2606.29971v1 📥 PDF

作者: Aydin Javadov, Shyngys Aitkazinov, Tobias Hoesli, Florian von Wangenheim, Bjoern Schuller, Joseph Ollier

分类: cs.LG

发布日期: 2026-06-29

💡 一句话要点

提出NeuReasoner以解决推理引发边界问题

🎯 匹配领域: 支柱九：具身大模型 (Embodied Foundation Models)

关键词: 推理能力 大型语言模型 认知心理学 模块化设计 理论驱动 行为任务 算术推理 代码生成

📋 核心要点

  1. 现有方法主要依赖数学和编码基准，未能全面探索大型语言模型推理能力的边界条件。
  2. NeuReasoner通过结合神经透镜和认知透镜，提供了一种新的引发推理的工具，旨在揭示推理能力的边界。
  3. 在CogBench等行为任务中，NeuReasoner在算术推理和代码生成等方面表现出显著提升，超越了传统推理基线。

📝 摘要（中文）

越来越多的研究表明，大型语言模型的推理能力在其基础形式中主要是潜在的，后续训练主要是放大而非引入这些能力。然而，这些证据主要来自数学和编码基准，导致这一主张的边界条件尚未得到充分探索。为此，本文提出了NeuReasoner，一个基于理论的引发工具。通过将神经透镜与认知透镜相结合，NeuReasoner在CogBench上进行评估，显示出在算术推理、代码生成等任务中超越传统推理的能力，并明确了推理引发的成功与失败边界。

🔬 方法详解

问题定义：本文旨在解决大型语言模型推理能力的边界条件问题，现有方法未能充分探索哪些认知任务可以通过引发获得，哪些则无法恢复。

核心思路：NeuReasoner的核心思路是结合神经透镜和认知透镜，通过模块化的方式整合输出，旨在通过理论驱动的方式引发推理能力的显现。

技术框架：NeuReasoner的整体架构包括两个主要模块：神经透镜和认知透镜。神经透镜负责功能特异性的引导，而认知透镜则基于推理的提问理论，二者的输出通过内部模块化整合。

关键创新：NeuReasoner的最大创新在于其理论驱动的引发机制，通过模块化设计实现了对推理能力的系统性探索，超越了以往仅依赖数学和编码基准的研究。

关键设计：在设计中，NeuReasoner采用了特定的参数设置和损失函数，以确保在不同认知任务中的有效性，同时保持模型的可解释性和模块化特征。具体的网络结构和参数设置在实验中进行了优化。

📊 实验亮点

NeuReasoner在CogBench的评估中显示出在算术推理、代码生成、贝叶斯推理和奖励学习等任务上，能够匹配或超越思维模式基线，尤其在自一致性和迭代精炼基线的对比中，表现出显著的提升，验证了其有效性。

🎯 应用场景

NeuReasoner的研究成果在多个领域具有潜在应用价值，包括教育、心理学和人工智能系统的推理能力提升。通过明确推理能力的边界，未来可以更好地设计和优化智能系统，使其在复杂认知任务中表现更佳。

📄 摘要（原文）

A growing body of work suggests that the reasoning capabilities of large language models are largely latent in their base form, with post-training primarily amplifying rather than introducing them. However, this evidence comes mainly from mathematical and coding benchmarks, leaving the boundary conditions of that claim largely unexplored, namely which cognitive tasks can be recovered through elicitation and where that recovery fails. To investigate this, we introduce NeuReasoner, a theory-grounded elicitation instrument. At each step, an orchestrator pairs a Neuro Lens, inspired by functional specificity, with a Cognitive Lens, drawn from the Erotetic Theory of Reasoning, and integrates their outputs through internal modularization of a single model, without external tools. We evaluate NeuReasoner on CogBench, a suite of behavioral tasks from cognitive psychology, alongside standard mathematical and coding benchmarks, measuring both its improvement over vanilla inference and its ability to match a model's post-trained thinking mode. At sufficient scale, NeuReasoner matches or exceeds thinking-mode baselines on arithmetic reasoning, code generation, Bayesian reasoning, and reward learning; these gains persist against self-consistency and iterative-refinement baselines matched to NeuReasoner's per-decision call budget. Using NeuReasoner allows us to find clear boundaries: risk-taking and decision making under uncertainty remains hard to recover through elicitation alone, and model scale interacts with elicitation in both directions: widening its advantage on some cognitive signatures while erasing it on others. Overall, through NeuReasoner as a modular, interpretable, theory-grounded elicitation instrument, we empirically map where reasoning elicitation succeeds and fails, beyond the mathematical and coding benchmarks where prior claims have rested.