Publications

Fifty-six papers,
CORE A* venues

Every entry below appeared at a CORE A*-ranked venue between 2016 and 2026. The full record of 273 publications is on Google Scholar and DBLP.

Selected

2011Tensor-Train DecompositionSIAM J. Sci. Comput. 4,100+ citations. The founding result.
2015Time Integration of Tensor TrainsSIAM J. Numer. Anal. SIAM Outstanding Paper Prize 2018.
2015Speeding-up Convolutional Neural Networks Using Fine-tuned CP-DecompositionICLR. Early tensor compression of deep nets.
2020Hyperbolic Image EmbeddingsCVPR. Geometry beyond Euclidean for vision.
2024Your Transformer is Secretly LinearACL. Near-linear maps between transformer layers.

All fifty-six

56 papers

2026

EACLConfidence Leaps in LLM Reasoning: Early Stopping and Cross-Model TransferPavel Tikhonov, Ivan V. Oseledets, and Elena Tutubalina. Conference of the European Chapter of the Association for Computational Linguistics (EACL).
FindingsFeature Drift: How Fine-Tuning Repurposes Representations in LLMsAndrey V. Galichin, Anton Korznikov, Alexey Dontsov, Oleg Rogov, Elena Tutubalina, and Ivan V. Oseledets. Findings of the Association for Computational Linguistics: EACL.
AAAII Have Covered All the Bases Here: Interpreting Reasoning Features in Large Language Models via Sparse AutoencodersAndrey V. Galichin, Alexey Dontsov, Polina Druzhinina, Anton Razzhigaev, Oleg Rogov, Elena Tutubalina, and Ivan V. Oseledets. AAAI Conference on Artificial Intelligence (AAAI).
WACVMaxInfo: A Training-Free Key-Frame Selection Method Using Maximum Volume for Enhanced Video UnderstandingPengyi Li, Irina Abdullaeva, Alexander Gambashidze, Andrey Kuznetsov, and Ivan V. Oseledets. IEEE/CVF Winter Conference on Applications of Computer Vision (WACV).
ACLOne Task Vector is not Enough: A Large-Scale Study for In-Context LearningPavel Tikhonov, Ivan V. Oseledets, and Elena Tutubalina. Annual Meeting of the Association for Computational Linguistics (ACL).

2025

IJCAIAI Diagnostic Assistant (AIDA): A Predictive Model for Diagnoses from Health Records in Clinical Decision Support SystemsDmitriy Umerenkov, Alexandr Nesterov, Vladimir Shaposhnikov, Ruslan Abramov, Nikolay Romanenko, Vladimir Kokh, Marina Kirina, Anton Abrosimov, Dmitry V. Dylov, and Ivan V. Oseledets. International Joint Conference on Artificial Intelligence (IJCAI).
ICLRAssociative memory and dead neuronsVladimir Fanaskov and Ivan V. Oseledets. International Conference on Learning Representations (ICLR).
AAAICertification of Speaker Recognition Models to Additive PerturbationsDmitrii Korzh, Elvir Karimov, Mikhail Pautov, Oleg Y. Rogov, and Ivan V. Oseledets. AAAI Conference on Artificial Intelligence (AAAI).
EMNLPCLARITY: Clinical Assistant for Routing, Inference, and TriageVladimir Shaposhnikov, Alexandr Nesterov, Ilia Kopanichuk, Ivan Bakulin, Egor Zhelvakov, Ruslan Abramov, Ekaterina Tsapieva, Iaroslav Bespalov, Dmitry V. Dylov, and Ivan V. Oseledets. Conference on Empirical Methods in Natural Language Processing (EMNLP): Industry Track.
FindingsCLEAR: Character Unlearning in Textual and Visual ModalitiesAlexey Dontsov, Dmitrii Korzh, Alexey Zhavoronkin, Boris Mikheev, Denis Bobkov, Aibek Alanov, Oleg Rogov, Ivan V. Oseledets, and Elena Tutubalina. Findings of the Association for Computational Linguistics: ACL.
EMNLPExploring the Hidden Capacity of LLMs for One-Step Text GenerationGleb Mezentsev and Ivan V. Oseledets. Conference on Empirical Methods in Natural Language Processing (EMNLP).
FindingsLLM-Microscope: Uncovering the Hidden Role of Punctuation in Context Memory of TransformersAnton Razzhigaev, Matvey Mikhalchuk, Temurbek Rahmatullaev, Elizaveta Goncharova, Polina Druzhinina, Ivan V. Oseledets, and Andrey Kuznetsov. Findings of the Association for Computational Linguistics: NAACL.
ACLRun LoRA Run: Faster and Lighter LoRA ImplementationsDaria Cherniuk, Aleksandr Mikhalev, and Ivan V. Oseledets. Annual Meeting of the Association for Computational Linguistics (ACL).

2024

ICMLNeural operators meet conjugate gradients: The FCG-NO method for efficient PDE solvingAlexander Rudikov, Vladimir Fanaskov, Ekaterina A. Muravleva, Yuri M. Laevsky, and Ivan V. Oseledets. International Conference on Machine Learning (ICML).
IJCAIProbabilistically Robust Watermarking of Neural NetworksMikhail Pautov, Nikita Bogdanov, Stanislav Pyatkin, Oleg Rogov, and Ivan V. Oseledets. International Joint Conference on Artificial Intelligence (IJCAI).
UAIQuantization of Large Language Models with an Overdetermined BasisDaniil Merkulov, Daria Cherniuk, Alexander Rudikov, Ivan V. Oseledets, Ekaterina A. Muravleva, Aleksandr Mikhalev, and Boris Kashin. Conference on Uncertainty in Artificial Intelligence (UAI).
CIKMRECE: Reduced Cross-Entropy Loss for Large-Catalogue Sequential RecommendersDanil Gusak, Gleb Mezentsev, Ivan V. Oseledets, and Evgeny Frolov. ACM International Conference on Information and Knowledge Management (CIKM).
RecSysScalable Cross-Entropy Loss for Sequential Recommendations with Large Item CatalogsGleb Mezentsev, Danil Gusak, Ivan V. Oseledets, and Evgeny Frolov. ACM Conference on Recommender Systems (RecSys).
RecSysSelf-Attentive Sequential Recommendations with Hyperbolic RepresentationsEvgeny Frolov, Tatyana Matveeva, Leyla Mirvakhabova, and Ivan V. Oseledets. ACM Conference on Recommender Systems (RecSys).
EMNLPSparseGrad: A Selective Method for Efficient Fine-tuning of MLP LayersViktoria Chekalina, Anna Rudenko, Gleb Mezentsev, Aleksandr Mikhalev, Alexander Panchenko, and Ivan V. Oseledets. Conference on Empirical Methods in Natural Language Processing (EMNLP).
FindingsThe Shape of Learning: Anisotropy and Intrinsic Dimensions in Transformer-Based ModelsAnton Razzhigaev, Matvey Mikhalchuk, Elizaveta Goncharova, Ivan V. Oseledets, Denis Dimitrov, and Andrey Kuznetsov. Findings of the Association for Computational Linguistics: EACL.
ACLYour Transformer is Secretly LinearAnton Razzhigaev, Matvey Mikhalchuk, Elizaveta Goncharova, Nikolai Gerasimenko, Ivan V. Oseledets, Denis Dimitrov, and Andrey Kuznetsov. Annual Meeting of the Association for Computational Linguistics (ACL).

2023

ICLRConstructive TT-representation of the tensors given as index interaction functions with applicationsGleb V. Ryzhakov and Ivan V. Oseledets. International Conference on Learning Representations (ICLR).
ICMLFew-bit Backward: Quantized Gradients of Activation Functions for Memory Footprint ReductionGeorgii Sergeevich Novikov, Daniel Bershatsky, Julia Gusak, Alex Shonenkov, Denis Valerievich Dimitrov, and Ivan V. Oseledets. International Conference on Machine Learning (ICML).
ICMLGeneral Covariance Data Augmentation for Neural PDE SolversVladimir Fanaskov, Tianchi Yu, Alexander Rudikov, and Ivan V. Oseledets. International Conference on Machine Learning (ICML).
FindingsLayerwise universal adversarial attack on NLP modelsOlga Tsymboi, Danil Malaev, Andrei Petrovskii, and Ivan V. Oseledets. Findings of the Association for Computational Linguistics: ACL.
NeurIPSNeural Harmonics: Bridging Spectral Embedding and Matrix Completion in Self-Supervised LearningMarina Munkhoeva and Ivan V. Oseledets. Neural Information Processing Systems (NeurIPS).
NeurIPSPROTES: Probabilistic Optimization with Tensor SamplingAnastasia Batsheva, Andrei Chertkov, Gleb V. Ryzhakov, and Ivan V. Oseledets. Neural Information Processing Systems (NeurIPS).
ICLRUnderstanding DDPM Latent Codes Through Optimal TransportValentin Khrulkov, Gleb V. Ryzhakov, Andrei Chertkov, and Ivan V. Oseledets. International Conference on Learning Representations (ICLR).

2022

AAAICC-CERT: A Probabilistic Approach to Certify General Robustness of Neural NetworksMikhail Pautov, Nurislam Tursynbek, Marina Munkhoeva, Nikita Muravev, Aleksandr Petiushko, and Ivan V. Oseledets. AAAI Conference on Artificial Intelligence (AAAI).
WACVGeometry-Inspired Top-k Adversarial PerturbationsNurislam Tursynbek, Aleksandr Petiushko, and Ivan V. Oseledets. IEEE/CVF Winter Conference on Applications of Computer Vision (WACV).
CVPRHyperbolic Vision Transformers: Combining Improvements in Metric LearningAleksandr Ermolov, Leyla Mirvakhabova, Valentin Khrulkov, Nicu Sebe, and Ivan V. Oseledets. IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR).
NeurIPSSmoothed Embeddings for Certified Few-Shot LearningMikhail Pautov, Olesya Kuznetsova, Nurislam Tursynbek, Aleksandr Petiushko, and Ivan V. Oseledets. Neural Information Processing Systems (NeurIPS).
IJCAISurvey on Efficient Training of Large Neural NetworksJulia Gusak, Daria Cherniuk, Alena Shilova, Alexandr Katrutsa, Daniel Bershatsky, Xunyi Zhao, Lionel Eyraud-Dubois, Oleh Shliazhko, Denis Dimitrov, Ivan V. Oseledets, and Olivier Beaumont. International Joint Conference on Artificial Intelligence (IJCAI).
NeurIPSTTOpt: A Maximum Volume Quantized Tensor Train-based Optimization and its Application to Reinforcement LearningKonstantin Sozykin, Andrei Chertkov, Roman Schutski, Anh-Huy Phan, Andrzej S. Cichocki, and Ivan V. Oseledets. Neural Information Processing Systems (NeurIPS).

2021

AAAIAdversarial Turing Patterns from Cellular AutomataNurislam Tursynbek, Ilya Vilkoviskiy, Maria Sindeeva, and Ivan V. Oseledets. AAAI Conference on Artificial Intelligence (AAAI).
ICMLFunctional Space Analysis of Local GAN ConvergenceValentin Khrulkov, Artem Babenko, and Ivan V. Oseledets. International Conference on Machine Learning (ICML).
ICCVLatent Transformations via NeuralODEs for GAN-based Image EditingValentin Khrulkov, Leyla Mirvakhabova, Ivan V. Oseledets, and Artem Babenko. IEEE/CVF International Conference on Computer Vision (ICCV).
UAITensor-train density estimationGeorgii S. Novikov, Maxim E. Panov, and Ivan V. Oseledets. Conference on Uncertainty in Artificial Intelligence (UAI).

2020

CVPRHyperbolic Image EmbeddingsValentin Khrulkov, Leyla Mirvakhabova, Evgeniya Ustinova, Ivan V. Oseledets, and Victor S. Lempitsky. IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR).
NeurIPSInterpolation Technique to Speed Up Gradients Propagation in Neural ODEsTalgat Daulbaev, Alexandr Katrutsa, Larisa Markeeva, Julia Gusak, Andrzej Cichocki, and Ivan V. Oseledets. Neural Information Processing Systems (NeurIPS).
RecSysPerformance of Hyperbolic Geometry Models on Top-N Recommendation TasksLeyla Mirvakhabova, Evgeny Frolov, Valentin Khrulkov, Ivan V. Oseledets, and Alexander Tuzhilin. ACM Conference on Recommender Systems (RecSys).
ECCVStable Low-Rank Tensor Decomposition for Compression of Convolutional Neural NetworkAnh Huy Phan, Konstantin Sobolev, Konstantin Sozykin, Dmitry Ermilov, Julia Gusak, Petr Tichavský, Valeriy Glukhov, Ivan V. Oseledets, and Andrzej Cichocki. European Conference on Computer Vision (ECCV).
FindingsTensorized Embedding LayersOleksii Hrinchuk, Valentin Khrulkov, Leyla Mirvakhabova, Elena Orlova, and Ivan V. Oseledets. Findings of the Association for Computational Linguistics: EMNLP.
ICLRThe Shape of Data: Intrinsic Distance for Data DistributionsAnton Tsitsulin, Marina Munkhoeva, Davide Mottin, Panagiotis Karras, Alexander M. Bronstein, Ivan V. Oseledets, and Emmanuel Müller. International Conference on Learning Representations (ICLR).

2019

ICLRGeneralized Tensor Models for Recurrent Neural NetworksValentin Khrulkov, Oleksii Hrinchuk, and Ivan V. Oseledets. International Conference on Learning Representations (ICLR).
RecSysHybridSVD: when collaborative information is not enoughEvgeny Frolov and Ivan V. Oseledets. ACM Conference on Recommender Systems (RecSys).
ICMLPROVEN: Verifying Robustness of Neural Networks with a Probabilistic ApproachLily Weng, Pin-Yu Chen, Lam M. Nguyen, Mark S. Squillante, Akhilan Boopathy, Ivan V. Oseledets, and Luca Daniel. International Conference on Machine Learning (ICML).

2018

CVPRArt of Singular Vectors and Universal Adversarial PerturbationsValentin Khrulkov and Ivan V. Oseledets. IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR).
ICLRExpressive power of recurrent neural networksValentin Khrulkov, Alexander Novikov, and Ivan V. Oseledets. International Conference on Learning Representations (ICLR).
ICMLGeometry Score: A Method For Comparing Generative Adversarial NetworksValentin Khrulkov and Ivan V. Oseledets. International Conference on Machine Learning (ICML).
NeurIPSQuadrature-based features for kernel approximationMarina Munkhoeva, Yermek Kapushev, Evgeny Burnaev, and Ivan V. Oseledets. Neural Information Processing Systems (NeurIPS).

2017

ICLRExponential MachinesAlexander Novikov, Mikhail Trofimov, and Ivan V. Oseledets. ICLR Workshop Track.
ACLRiemannian Optimization for Skip-Gram Negative SamplingAlexander Fonarev, Oleksii Hrinchuk, Gleb Gusev, Pavel Serdyukov, and Ivan V. Oseledets. Annual Meeting of the Association for Computational Linguistics (ACL).

2016

ICDMEfficient Rectangular Maximal-Volume Algorithm for Rating Elicitation in Collaborative FilteringAlexander Fonarev, Alexander Mikhalev, Pavel Serdyukov, Gleb Gusev, and Ivan V. Oseledets. Proceedings of ICDM.
RecSysFifty Shades of Ratings: How to Benefit from a Negative Feedback in Top-N Recommendations TasksEvgeny Frolov and Ivan V. Oseledets. ACM Conference on Recommender Systems (RecSys).

Google Scholar: all 273 publications →DBLP author record →

Fifty-six papers,CORE A* venues

2026

2025

2024

2023

2022

2021

2020

2019

2018

2017

2016

Fifty-six papers,
CORE A* venues