Learning Using Privileged Information

Overview

The field studies how extra information available during training—such as future observations, detailed labels, other sensor modalities, or expert feedback—can help a model that will not receive it at deployment. Common approaches include teacher–student distillation, privileged-feature SVMs, representation learning, and reinforcement learning with guided or hindsight-informed rollouts. These methods are used in settings from robotics and autonomous driving to medical imaging, vision-language models, and reasoning agents. The central tension is that richer guidance can improve learning, but may also teach shortcuts or bias the student toward evidence it cannot access, so useful supervision must be relevant, reliable, and transferable to ordinary inputs.

Latest papers

CardsList

76 papers8 papers in the last four weeks, down 50% on the four weeks before. 0.1% of all new papers.

  1. RAIL: Representation-Aligned Imitation Learning for Student-Compatible Teacher Policies

    Nov 9, 2026Meraj Mammadov, Pedro Zuidberg Dos Martires, Johannes Andreas StorkImitation LearningRL for Knowledge Distillation

  2. Not Yet in the Profile: Uncertainty-Aware Dynamic User Profiling by Learning from Future Interactions

    Nov 8, 2026Jieyong Kim, Neil Shah, Dong WangUser ModelingUser Preference Modeling

  3. TouchAnnotator: Multi-Source Tactile Supervision with Privileged-View Distillation

    Nov 6, 2026Ao Li, Zhihao Yuan, Chubin Zhang +6Visual-to-Tactile PredictionVisuo-Tactile Representation Learning

  4. Aligning Policy Encoders with a Shape Teacher

    Nov 5, 2026Madelon Bernardy, Mohammad Mahdi Derakhshani, David M Knigge +2Visuomotor Policy LearningRobotic Manipulation

  5. EvoOPD: Evolving Privileged Information Strategies for Black-Box On-Policy Distillation

    Nov 4, 2026Ximo Zhu, Ruiqi Liu, Yi Han +6Black-Box Knowledge DistillationLanguage Model Distillation

  6. Learning Fine-Grained Visual Classification with Text as Privileged Information via Multimodal Distillation

    Oct 23, 2026Aishwarya Venkataramanan, Christian Ickler, Joachim DenzlerFine-Grained Image ClassificationLearning Using Privileged Information

  7. Response-Informed Value Estimation for Cost-Aware Multi-Turn LLM Routing

    Oct 16, 2026Haotian Yan, Jiaqi Liu, Bingyi Jing +1LLM RoutingCost-Aware Inference

  8. Predicting Privileged Information Effectiveness for On-Policy Self-Distillation

    Oct 15, 2026Kaijian Zou, Emiliano Penaloza, MohammadHossein Rezaei +5On-Policy Self-DistillationLearning Using Privileged Information

  9. Motion- and Gravity-Aware Feedforward 3D Reconstruction

    Oct 15, 2026Shiyao Qian, Yuan Ren, Dongfeng BaiMulti-View 3D ReconstructionFeed-Forward 3D Reconstruction

  10. Learning Individual Scientific Taste with Privileged Hindsight Distillation

    Oct 14, 2026Hanwei Fan, Yanqiao Wang, Yiqi Zhu +3AI-Assisted Scientific ResearchLearning Using Privileged Information

  11. Not Every Cue Is Evidence: Cue Supportability for Forensic Distillation in Face Forgery Detection

    Oct 12, 2026Binjia Zhou, Dawei Luo, Shuai Chen +7Explainable Deepfake DetectionDeepfake Detection

  12. Self-Distillation with Verbal Regularization

    Oct 12, 2026Jiazheng Li, Yawei Li, Junru Lu +7On-Policy Self-DistillationLanguage Model Self-Improvement

  13. FPD-Drive: Future Policy Distillation for End-to-End Autonomous Driving

    Oct 9, 2026Li-Heng Chen, Qifeng Chen, Chengwei Wu +4End-to-End Autonomous DrivingAutonomous Driving Planning

  14. Seeing Through the Displaced Frame: Privileged Noise Distillation for Vision-Force Precision Assembly

    Oct 6, 2026Ching-Hsiang Chang, Tzu-Yu Chuang, Yi-Hsiu Lee +3Robotic ManipulationForce-Controlled Robotic Manipulation

  15. Masked Privileged-Information Distillation for Multimodal Skin Lesion Classification Under Missing Clinical Metadata

    Oct 2, 2026Anirban Barua, Md Mahir Abrar Khan, Ayman Iktidar +1Skin Lesion ClassificationMultimodal Classification

  16. ReSAIL: Mitigating Collapse in Iterative Agent Self-Distillation

    Sep 30, 2026Shengjie Jin, Hengbo Xu, Zelong Sun +2On-Policy Self-DistillationLLM Agent Self-Improvement

  17. RAST: Resolution-Aware Privileged Structure Transfer for Low-Resolution Audio Activity Recognition

    Sep 30, 2026Ji Hwan Park, Gautham Krishna Gudur, Yufei Shen +2Audio Representation LearningHuman Activity Recognition

  18. Instruct, Not Answer: Using Instruction Privileges in On-Policy Context Distillation

    Sep 26, 2026Hantao Yu, Xiaoxue Han, Udaya Ghai +4OOD GeneralizationOn-Policy Distillation

  19. From Self-Distillation to Self-Practice: Privileged Information for Multi-Turn Agents

    Sep 24, 2026Xingyu Su, Abhishek Kumar, Qing Ping +5On-Policy Self-DistillationLLM Agent Training

  20. PLAT: Sparse Timed Keyframe Motion Tracking for Humanoid Control via Privileged Latent Transition Learning

    Sep 22, 2026Zepeng Wang, Jiangxing Wang, Chao Ma +2Robot Imitation LearningHumanoid Motion Tracking

  21. What Should a Self-Teacher See? Privileged Context Design for On-Policy Self-Distillation

    Sep 22, 2026Kanghui Tian, Siyuan Liu, Tianxiang Jiang +8On-Policy Self-DistillationContext Distillation

  22. Mind2Dialogue: Training Human-Aware Language Models by Simulating User Mental States

    Sep 14, 2026Zixuan Wang, Yufan Zhou, Jinzhou Tang +16LLM PersonalizationTheory of Mind

  23. Quantifying the Value of Privileged Information Using a PAC-Bayesian Approach

    Sep 11, 2026Vasily Bokov, Sebastian Schmitt, Vedran Dunjko +1Value of InformationLearning Using Privileged Information

  24. SCOPE-OPSD: Fisher-Conditioned Privileged Subspaces for On-Policy Self-Distillation

    Sep 11, 2026Yunmeng Chen, Kunyu Wang, Peihan Li +8On-Policy Self-DistillationLanguage Model Distillation

  25. The microscope is the mask: privileged views and labels from a cryo-ET forward model

    Sep 3, 2026Bogdan Toader, Kiarash Jamali, Tanmay A. M. Bharat +1Protein Representation LearningCryo-Electron Microscopy

  26. Learning to Track from Privileged Target Appearances

    Sep 2, 2026Xin Chen, Jiao Xu, Dong Wang +2Visual Object TrackingLearning Using Privileged Information