Representation Drift

Ask about this topic

Overview

Representation drift studies how a model’s internal features reorganize as it learns new tasks, adapts to new data, or undergoes fine-tuning, and how those changes relate to forgetting, safety, and performance. Researchers measure drift with layer-wise activation comparisons, geometric and spectral analyses of feature spaces, and sparse-feature diagnostics, often testing them on continual-learning, retrieval, and multimodal tasks. Methods such as replay, low-rank adaptation, representation regularization, and memory-based updates aim to balance learning new information with preserving useful features. A central tension is that drift can signal harmful forgetting or safety failures, but it can also support useful adaptation; its size alone does not establish whether a change is beneficial or causally responsible for behavior.

Latest papers

CardsList

20 papers5 papers in the last four weeks, against 2 the four weeks before. 0.0% of all new papers.

  1. Sparse Autoencoders Can Induce Local Geometry for Mitigating Catastrophic Forgetting

    Oct 15, 2026Kamil Faber, Katarzyna Filus, Łukasz PawelaCatastrophic ForgettingRegularization-Based Continual Learning

    Faculty of Computer Science, AGH University of Krakow, Krakow, Poland · Institute of Theoretical and Applied Informatics, Polish Academy of Sciences, Gliwice, Poland

  2. Not All Drift Is a Mechanism: Tracing Prompt Adaptation in LLMs

    Oct 13, 2026Mikhail Aleksandrov, Nail Bashirov, Maksim Ivanov +6Prefix TuningLLM Prompting

    BRAIn Lab

  3. Representation Drift in Continual Learning: LoRA versus N-gram Memory

    Oct 10, 2026Hao WuContinual Learning for LLMsOnline Continual Learning

    Illinois Institute of Technology

  4. Quantize by Drift: Label-Free Mixed-Precision Post-Training Quantization for Text Embedders

    Oct 6, 2026Hyojung Han, Jongmin Kim, Seung-Hun JeonPost-Training QuantizationMixed-Precision Quantization

    ThakiCloud

  5. Early Learning Shapes Later Directions Of Representation Change In Continual Learning

    Sep 28, 2026Yuantao Deng, Jinnuo Liu, Kaizhen Tan +1Neural Representation GeometryVisual Representation Learning

    New York University · Xi’an Jiaotong University

  6. Reuse or Relearn? A Spectral View of Earth Observation Foundation Models

    Sep 26, 2026Mehmet Ozgur Turkoglu, Valerio Marsocci, Dominik J. Mühlematter +3Fine-TuningGeospatial Foundation Models

    Agroscope, EO of Agroecosystems · ESA, Φ-lab · ETH Zurich

  7. Capsule Lens: Locating and Tracking Concept Geometry in Model Representations

    Sep 4, 2026Yiming Tang, Harshvardhan Saini, Samyak Jha +3Neural Network InterpretabilityNeural Representation Geometry

    Indian Institute of Technology (ISM) Dhanbad · The Chinese University of Hong Kong · National University of Singapore

  8. Geometry of Forgetting: Representation Flux in Continual Learning

    Aug 16, 2026Maksim A. KazanskiiCatastrophic ForgettingRepresentation Drift

  9. Drift and Dependence: Layer-wise Information-Theoretic Bounds for Replay-Based Continual Learning

    Aug 12, 2026Tieliang Gong, Zhongbo Zhang, Wen Wen +1Information-Theoretic Generalization BoundsReplay-Based Continual Learning

    Tieliang Gong, Zhongbo Zhang and Wen Wen are with the School of Computer Science and Technology, Xi’an Jiaotong University, China · Yong-Jin Liu is with Department of Computer Science and Technology, Tsinghua University, China

  10. Do Activation Monitors Survive Model Updates? Benchmarking, Predicting, and Repairing Activation-Monitor Staleness

    Jun 14, 2026Evan DuanLanguage Model Safety EvaluationRepresentation Drift

    University of Michigan

  11. The Stable Recovery Manifold: Geometric Principles Governing Recoverability in Continual Learning

    Jun 11, 2026Ayushman Trivedi, Bhavika MelwaniContinual LearningGeometric Representation Learning

  12. Why Can't They Remember? Uncovering Representation and Retrieval Bottlenecks in Multi-Turn Acoustic Memory

    May 26, 2026Yang Xiao, Siyi Wang, Han Yin +4Audio-Language Model EvaluationMemorization in Language Models

    1The University of Melbourne · 2KAIST · 3The University of Auckland +1

  13. Safety Geometry Collapse in Multimodal LLMs and Adaptive Drift Correction

    May 18, 2026Jiahe Guo, Xiangran Guo, Jiaxuan Chen +6Multimodal Large Language ModelsLLM Safety

    Harbin Institute of Technology · Huawei Technologies Co., Ltd

  14. A Mechanistic Investigation of Supervised Fine Tuning

    May 12, 2026Ruhaan ChopraSupervised Fine-TuningLLM Alignment

    Independent Researcher

  15. SR2^2-LoRA: Self-Rectifying Inter-layer Relations in Low-Rank Adaptation for Class-Incremental Learning

    May 8, 2026Fengqiang Wan, Yipeng Lin, Kan Lv +1Representation DriftLow-Rank Adaptation

    School of Computer Science and Engineering, Nanjing University of Science and Technology, Nanjing 210094, China

  16. The Geometric Canary: Predicting Steerability and Detecting Drift via Representational Stability

    Apr 20, 2026Prashant C. RajuLanguage Model SteeringLLM Reliability

  17. Decomposing the Depth Profile of Fine-Tuning

    Apr 19, 2026Jayadev BillaFine-TuningRepresentation Drift

  18. Continual Adaptation for Pacific Indigenous Speech Recognition

    Mar 6, 2026Yang Xiao, Aso Mahmudi, Nick Thieberger +3Automatic Speech RecognitionLow-Resource Speech Recognition

    The University of Melbourne · UNSW Sydney

  19. Representation Stability in a Minimal Continual Learning Agent

    Feb 23, 2026Vishnu SubramanianContinual LearningLifelong Learning Agents

    AI/ML Computational Science Associate Manager, Accenture

  20. When Backdoors Go Beyond Triggers: Semantic Drift in Diffusion Models Under Encoder Attacks

    Feb 21, 2026Shenyang Chen, Liuwan ZhuAdversarial AttacksRepresentation Drift

    Google · Electrical and Computer Engineering Department, University of Hawai‘i at Mānoa