Large language models frequently produce errors in reasoning tasks despite possessing the underlying knowledge required for correct reasoning. One possible approach to improve reasoning consistency is through activation steering. However, existing activation steering approaches apply fixed, pre-computed correction vectors, ignoring where the model currently sits along its generation trajectory; the result is indiscriminate perturbation that disrupts already-correct steps as freely as erroneous ones. We propose Manifold-Guided Attention Steering (MAGS), a trajectory-aware inference-time intervention grounded in a geometric observation: the output activations of specific attention heads diverge from a low-dimensional correctness manifold at the point of error, and this deviation compounds through subsequent steps. For each identified attention head, we learn a low-dimensional subspace from contrastive pairs of correct and incorrect traces that capture the directions along which error behavior deviates from correct behavior. During inference, we monitor each head's proximity to this manifold and apply a targeted projection correction when deviation exceeds a learned threshold, steering the attention output back toward the correct subspace before the error propagates. MAGS consistently outperforms both unsteered baselines and static steering approaches across benchmarks spanning mathematical reasoning (MATH-500, GSM8K), code generation (HumanEval, MBPP), and molecular generation (SMILES), suggesting that correctness manifolds are a general feature of LLM attention geometry.
Interpretable Steering of Large Language Models with Feature Guided Activation Additions
Raffles Science Institute(莱佛士科学学院);Raffles Institution(莱佛士书院);Nous Research(努斯研究机构);Centre for Frontier AI Research (CFAR)(前沿人工智能研究中心);Agency for Science Technology and Research (A*STAR)(新加坡科技研究局)
Euphonium: Steering Video Flow Matching via Process Reward Gradient Guided Stochastic Dynamics
Huazhong University of Science and Technology(华中科技大学);Shanghai Jiao Tong University(上海交通大学);Tencent Hunyuan(腾讯混元);University of Chinese Academy of Sciences(中国科学院大学)
Mitigating Overthinking in Large Reasoning Models via Manifold Steering
Institute of Artificial Intelligence, Beihang University(北京航空航天大学人工智能研究院);College of AI, Tsinghua University(清华大学人工智能学院);Shanghai Qi Zhi Institute(上海启智研究所);State Key Laboratory of Virtual Reality Technology and Systems, Beihang University(北京航空航天大学虚拟现实技术与系统国家重点实验室)
Locate-then-Sparsify: Attribution Guided Sparse Strategy for Visual Hallucination Mitigation
State Key Lab. of AI Safety, Institute of Computing Technology, Chinese Academy of Sciences(中国科学院人工智能安全国家重点实验室,计算技术研究所);School of Advanced Interdisciplinary Sciences, University of Chinese Academy of Sciences(中国科学院大学先进交叉科学学院);School of Computer Science and Technology, University of Chinese Academy of Sciences(中国科学院大学计算机科学与技术学院)
本文提出了一种名为Locate-Then-Sparsify for Feature Steering (LTS-FS)的框架,通过定位和稀疏化策略,根据每层与幻觉的相关性调整特征引导强度,从而有效缓解视觉语言模型中的幻觉问题,同时保持良好的性能。
Forest-Guided Semantic Transport for Label-Supervised Manifold Alignment
Department of Mathematics and Statistics, Université de Montréal, Montreal, Canada(蒙特利尔大学数学与统计学系);Mila - Quebec AI Institute, Montreal, Canada(魁北克人工智能研究所);Department of Mathematics and Statistics, Utah State University, Logan, Utah, USA(犹他州立大学数学与统计学系);Department of Statistics, Brigham Young University, Provo, Utah, USA(Brigham Young 大学统计学系)
BarrierSteer: LLM Safety via Learning Barrier Steering
Department of Computer Science, National University of Singapore(新加坡国立大学计算机科学系);Singapore-MIT Alliance for Research and Technology Centre(新加坡-麻省理工联合研究中心);CSAIL, Massachusetts Institute of Technology(麻省理工学院计算机科学与人工智能实验室);Worcester Polytechnic Institute(沃斯堡理工学院)
Guided Manifold Alignment with Geometry-Regularized Twin Autoencoders
Department of Statistics(统计学系);Department of Computer Science(计算机科学系);Neuroscience Center(神经科学中心);Neuroscience Center and Department of Psychology(神经科学中心和心理学系);Department of Psychology(心理学系)