pith. sign in

arxiv: 2607.01918 · v1 · pith:RALWHSO2new · submitted 2026-07-02 · 💻 cs.LG

Zeus: Towards Tuning-Free Foundation Model for Time Series Analysis

Yisong Fu , Zezhi Shao , Chengqing Yu , Yujie Li , Yongjun Xu , Xueqi Cheng , Fei Wang This is my paper

Pith reviewed 2026-07-03 17:32 UTC · model grok-4.3

classification 💻 cs.LG
keywords time series foundation modeltuning-freemulti-scale transformermulti-objective temporal maskingpoint-wise tokenizationtime series analysiszero-shot generalization
0
0 comments X

The pith

Zeus delivers competitive results on five time series tasks without any task-specific fine-tuning.

A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.

The paper introduces Zeus, a single foundation model for time series that aims to eliminate the usual requirement for task-by-task fine-tuning. It tackles two core obstacles to multi-task generalization: balancing fine point-level detail against the cost of long sequences, and handling the different biases needed for tasks such as forecasting versus imputation. The authors argue that a multi-scale Transformer with point-wise tokenization plus a U-shaped hierarchy, combined with a single Multi-Objective Temporal Masking scheme, is sufficient to meet both requirements. If the approach holds, practitioners could apply one pretrained model across heterogeneous time-series problems without additional training steps.

Core claim

Zeus is a unified tuning-free Time Series Foundation Model that reconciles point-level granularity with long-sequence scalability through a multi-scale Transformer that uses point-wise tokenization and a U-shaped hierarchy, while Multi-Objective Temporal Masking accommodates the distinct inductive biases of extrapolation, interpolation, and global abstraction tasks inside one training regime, yielding competitive performance across five representative tasks in a fully tuning-free setting.

What carries the argument

Multi-scale Transformer with point-wise tokenization and U-shaped hierarchy, together with Multi-Objective Temporal Masking (MOTM)

If this is right

  • A single pretrained model can be applied directly to extrapolation, interpolation, and abstraction tasks without separate adaptation steps.
  • Point-level predictions remain feasible even when input sequences are long.
  • Heterogeneous task biases are handled by one masking objective rather than multiple specialized heads or losses.
  • Computational overhead of repeated fine-tuning across tasks is avoided.

Where Pith is reading between the lines

These are editorial extensions of the paper, not claims the author makes directly.

  • If the architecture generalizes, similar multi-scale plus multi-objective masking designs could be tested on other sequential domains such as audio or video.
  • The approach suggests that foundation-model scale may reduce the traditional need for per-task hyperparameter search in time-series work.
  • Longer sequences or streaming settings could serve as a direct test of whether the U-shaped hierarchy continues to control memory and compute costs.

Load-bearing premise

The combination of point-wise tokenization, U-shaped multi-scale hierarchy, and Multi-Objective Temporal Masking is enough to remove any need for task-specific fine-tuning while preserving accuracy on point-level and long-sequence problems.

What would settle it

A controlled comparison in which, on any one of the five tasks, a model that receives task-specific fine-tuning produces statistically higher accuracy than Zeus in its tuning-free configuration would falsify the central claim.

Figures

Figures reproduced from arXiv: 2607.01918 by Chengqing Yu, Fei Wang, Xueqi Cheng, Yisong Fu, Yongjun Xu, Yujie Li, Zezhi Shao.

Figure 1
Figure 1. Figure 1: Overall performance comparison of ZEUS under the tuning-free setting. ZEUS surpasses full-shot task-specific mod￾els (dashed lines) and significantly outperforms other TSFMs in tuning-free setting (solid lines). Inspired by the success of foundation models in language (OpenAI, 2023), images (Ramesh et al., 2021), and videos (Liu et al., 2024c), researchers have been striving to develop general-purpose time… view at source ↗
Figure 2
Figure 2. Figure 2: Overall architecture of ZEUS. Inputs from different downstream tasks are first unified into a common format and converted into point-wise tokens via tokenization. The resulting sequence is then processed by a U-shaped multi-scale Transformer. Quantile head is used to produce probabilistic outputs, while for classification tasks, global pooling is applied to obtain sequence-level representations. recovering… view at source ↗
Figure 3
Figure 3. Figure 3: The MOTM pipeline. MOTM hierarchically determines the masking ratio, scales the temporal scope, and applies diverse masking strategies to jointly optimize for extrapolation, interpola￾tion, and local-global feature extraction. anomalies necessitate modeling global consistency (Liu et al., 2025a). Moreover, classification calls for both global abstraction and the identification of local shapelets (Le et al.… view at source ↗
Figure 5
Figure 5. Figure 5: Averaged accuracy on 26 UEA classification datasets, where LP denotes linear probing and prompt denotes fine-tuning on prompt tokens. See [PITH_FULL_IMAGE:figures/full_fig_p007_5.png] view at source ↗
Figure 6
Figure 6. Figure 6: Ablation results of ZEUS. probing (Goswami et al., 2024) for classification tasks. In our evaluation, ZEUS is primarily assessed in a tuning-free setting using a non-parametric 1-nearest neighbor (1-NN) classifier for evaluation, with optional PCA whitening ap￾plied for feature normalization. In addition, we also report linear probing results to assess the linear separability of the learned representations… view at source ↗
Figure 7
Figure 7. Figure 7: Multi-scale feature-norm heatmaps, which illustrates the roles of different scales: fine-scale representations are sensitive to local variations and extreme values (red boxes), mid-scale stripe patterns capture intrinsic periodicity, and coarse-scale representations model global pattern shifts (white vertical line) and contextual anomalies (yellow boxes). When a specific mask is removed, we keep the expect… view at source ↗
Figure 8
Figure 8. Figure 8: Efficiency comparison between ZEUS and Time-MoEbase, two point-tokenized models with comparable model size. Results are averaged over 1,000 runs on sequences of length L=4096. yellow boxes demonstrate that large-scale representations are effective in capturing contextual anomalies. Efficiency Analysis Conventional point-tokenized Trans￾formers suffers from the high computational cost, as pro￾cessing a sequ… view at source ↗
Figure 9
Figure 9. Figure 9: PMF of the geometric distribution used to sample missing segment lengths. The expected block length is 8, and with 99% probability the block length is smaller than 35. D.4. Anomaly Detection Benchmarks We evaluate the anomaly detection task on the UCR Anomaly Archive (Wu & Keogh, 2021), which consists of 250 tasks spanning diverse domains such as medicine, sports, entomology, and space science. The dataset… view at source ↗
Figure 10
Figure 10. Figure 10: Example of forecasts from ZEUS. 28 [PITH_FULL_IMAGE:figures/full_fig_p028_10.png] view at source ↗
Figure 11
Figure 11. Figure 11: Zero-shot examples of reconstruction by ZEUS. Blue boxes denote the anomalies identified through reconstruction. PenDigits (t-SNE) PenDigits (PCA) EigenWorms (t-SNE) Libras (t-SNE) EigenWorms (PCA) Libras (PCA) Cricket (t-SNE) Cricket (PCA) [PITH_FULL_IMAGE:figures/full_fig_p029_11.png] view at source ↗
Figure 12
Figure 12. Figure 12: Visualization of representations learned by ZEUS on the UEA datasets. 29 [PITH_FULL_IMAGE:figures/full_fig_p029_12.png] view at source ↗
read the original abstract

We present Zeus, a unified tuning-free Time Series Foundation Model (TSFM) that delivers superior performance across diverse analysis tasks without any task-specific fine-tuning. Unlike prior studies that primarily focus on zero-shot forecasting but require task-specific tuning for other tasks, Zeus bridges this gap by addressing two fundamental challenges in multi-task generalization. First, to reconcile point-level granularity with long-sequence scalability, Zeus incorporates a multi-scale Transformer featuring point-wise tokenization and a U-shaped hierarchy, effectively balancing fine-grained fidelity with computational efficiency. Second, to accommodate varying inductive biases across different tasks, Zeus introduces Multi-Objective Temporal Masking (MOTM), a unified strategy that supports heterogeneous tasks (e.g., extrapolation, interpolation, and global abstraction) within a single framework. Extensive experiments across five representative tasks demonstrate that Zeus consistently achieves competitive results in tuning-free settings, underscoring its potential as a general-purpose TSFM.

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, simulated authors' rebuttal, and a circularity audit. Tearing a paper down is the easy half of reading it; the pith above is the substance, this is the friction.

Referee Report

0 major / 2 minor

Summary. The paper introduces Zeus, a unified tuning-free Time Series Foundation Model (TSFM) that uses a multi-scale Transformer with point-wise tokenization and U-shaped hierarchy to balance point-level granularity and long-sequence scalability, together with Multi-Objective Temporal Masking (MOTM) to accommodate heterogeneous inductive biases across tasks such as extrapolation, interpolation, and global abstraction. It claims that this design enables competitive performance across five representative tasks without any task-specific fine-tuning.

Significance. If the reported results hold, the work would be a meaningful step toward general-purpose TSFMs that eliminate per-task tuning, addressing longstanding tensions between granularity, scalability, and task-specific biases in time series modeling. The explicit design rationale for reconciling these elements is a strength.

minor comments (2)
  1. Abstract: while the summary of the two core challenges and proposed solutions is clear, the abstract would be strengthened by naming the five tasks and reporting at least one key quantitative comparison (e.g., average rank or relative error) to make the central performance claim more concrete for readers.
  2. The manuscript would benefit from an explicit statement of the datasets used and the precise definition of 'tuning-free' (e.g., whether any hyper-parameters are still selected on a validation split) to allow direct replication of the claimed setting.

Simulated Author's Rebuttal

0 responses · 0 unresolved

We thank the referee for their positive assessment of our manuscript on Zeus, a tuning-free Time Series Foundation Model. The referee accurately summarizes the key contributions, including the multi-scale Transformer with point-wise tokenization and U-shaped hierarchy, as well as Multi-Objective Temporal Masking (MOTM) for handling diverse tasks. We appreciate the recognition of the work's potential significance toward general-purpose TSFMs and the recommendation for minor revision. No specific major comments were provided in the report.

Circularity Check

0 steps flagged

No significant circularity detected

full rationale

The paper introduces an architectural design (multi-scale Transformer with point-wise tokenization, U-shaped hierarchy, and MOTM) and reports empirical results on five tasks. No equations, derivations, predictions, or first-principles claims appear that could reduce by construction to fitted parameters, self-citations, or renamed inputs. The central claims rest on experimental validation rather than any load-bearing self-referential step, making the work self-contained against external benchmarks.

Axiom & Free-Parameter Ledger

0 free parameters · 0 axioms · 0 invented entities

Abstract-only review supplies no explicit free parameters, axioms, or invented entities; the two named components (multi-scale Transformer, MOTM) are presented as engineering choices rather than new postulates.

pith-pipeline@v0.9.1-grok · 5699 in / 1069 out tokens · 30181 ms · 2026-07-03T17:32:24.225063+00:00 · methodology

discussion (0)

Sign in with ORCID, Apple, or X to comment. Anyone can read and Pith papers without signing in.

Reference graph

Works this paper leans on

242 extracted references · 75 canonical work pages · 33 internal anchors

  1. [1]

    Scaling Learning Algorithms Towards

    Bengio, Yoshua and LeCun, Yann , booktitle =. Scaling Learning Algorithms Towards

  2. [2]

    and Osindero, Simon and Teh, Yee Whye , journal =

    Hinton, Geoffrey E. and Osindero, Simon and Teh, Yee Whye , journal =. A Fast Learning Algorithm for Deep Belief Nets , volume =

  3. [3]

    2016 , publisher=

    Deep learning , author=. 2016 , publisher=

    2016

  4. [4]

    Langley , title =

    P. Langley , title =. Proceedings of the 17th International Conference on Machine Learning (ICML 2000) , address =. 2000 , pages =

    2000

  5. [5]

    T. M. Mitchell. The Need for Biases in Learning Generalizations. 1980

    1980

  6. [6]

    M. J. Kearns , title =

  7. [7]

    Machine Learning: An Artificial Intelligence Approach, Vol. I. 1983

    1983

  8. [8]

    R. O. Duda and P. E. Hart and D. G. Stork. Pattern Classification. 2000

    2000

  9. [9]

    Suppressed for Anonymity , author=

  10. [10]

    Newell and P

    A. Newell and P. S. Rosenbloom. Mechanisms of Skill Acquisition and the Law of Practice. Cognitive Skills and Their Acquisition. 1981

    1981

  11. [11]

    A. L. Samuel. Some Studies in Machine Learning Using the Game of Checkers. IBM Journal of Research and Development. 1959

    1959

  12. [12]

    arXiv preprint arXiv:2308.08469 , year=

    Llm4ts: Two-stage fine-tuning for time-series forecasting with pre-trained llms , author=. arXiv preprint arXiv:2308.08469 , year=

    work page arXiv

  13. [13]

    arXiv preprint arXiv:2302.11939 , year=

    One Fits All: Power General Time Series Analysis by Pretrained LM , author=. arXiv preprint arXiv:2302.11939 , year=

    work page arXiv

  14. [14]

    arXiv preprint arXiv:2310.09751 , year=

    UniTime: A Language-Empowered Unified Model for Cross-Domain Time Series Forecasting , author=. arXiv preprint arXiv:2310.09751 , year=

    work page arXiv

  15. [15]

    2024 , title =

    Liu, Yong and Hu, Tengge and Zhang, Haoran and Wu, Haixu and Wang, Shiyu and Ma, Lintao and Long, Mingsheng , booktitle =. 2024 , title =

    2024

  16. [16]

    and Sinthong, Phanwadee and Kalagnanam, Jayant , booktitle =

    Nie, Yuqi and Nguyen, Nam H. and Sinthong, Phanwadee and Kalagnanam, Jayant , booktitle =. 2023 , title =

    2023

  17. [17]

    Proceedings of the AAAI conference on artificial intelligence , volume=

    Are transformers effective for time series forecasting? , author=. Proceedings of the AAAI conference on artificial intelligence , volume=

  18. [18]

    2023 , title =

    Wu, Haixu and Hu, Tengge and Liu, Yong and Zhou, Hang and Wang, Jianmin and Long, Mingsheng , booktitle =. 2023 , title =

    2023

  19. [19]

    2024 , title =

    donghao, Luo and xue, wang , booktitle =. 2024 , title =

    2024

  20. [20]

    arXiv preprint arXiv:2305.18803 , year=

    Koopa: Learning Non-stationary Time Series Dynamics with Koopman Predictors , author=. arXiv preprint arXiv:2305.18803 , year=

    work page arXiv

  21. [21]

    arXiv preprint arXiv:1905.10437 , year=

    N-BEATS: Neural basis expansion analysis for interpretable time series forecasting , author=. arXiv preprint arXiv:1905.10437 , year=

    work page arXiv 1905

  22. [22]

    Proceedings of the AAAI Conference on Artificial Intelligence , volume=

    Nhits: Neural hierarchical interpolation for time series forecasting , author=. Proceedings of the AAAI Conference on Artificial Intelligence , volume=

  23. [23]

    International Journal of Forecasting , volume=

    The M4 Competition: 100,000 time series and 61 forecasting methods , author=. International Journal of Forecasting , volume=. 2020 , publisher=

    2020

  24. [24]

    arXiv preprint arXiv:2210.03675 , year=

    Koopman neural forecaster for time series with temporal distribution shifts , author=. arXiv preprint arXiv:2210.03675 , year=

    work page arXiv

  25. [25]

    Advances in Neural Information Processing Systems , volume=

    Film: Frequency improved legendre memory model for long-term time series forecasting , author=. Advances in Neural Information Processing Systems , volume=

  26. [26]

    Advances in Neural Information Processing Systems , volume=

    Autoformer: Decomposition transformers with auto-correlation for long-term series forecasting , author=. Advances in Neural Information Processing Systems , volume=

  27. [27]

    arXiv preprint arXiv:2308.08241 , year=

    TEST: Text prototype aligned embedding to activate LLM's ability for time series , author=. arXiv preprint arXiv:2308.08241 , year=

    work page arXiv

  28. [28]

    arXiv preprint arXiv:2310.04948 , year=

    TEMPO: Prompt-based Generative Pre-trained Transformer for Time Series Forecasting , author=. arXiv preprint arXiv:2310.04948 , year=

    work page arXiv

  29. [29]

    IEEE Transactions on Knowledge and Data Engineering , year=

    Promptcast: A new prompt-based learning paradigm for time series forecasting , author=. IEEE Transactions on Knowledge and Data Engineering , year=

  30. [30]

    Neurocomputing , volume=

    Roformer: Enhanced transformer with rotary position embedding , author=. Neurocomputing , volume=. 2024 , publisher=

    2024

  31. [31]

    The Eleventh International Conference on Learning Representations , year=

    Crossformer: Transformer utilizing cross-dimension dependency for multivariate time series forecasting , author=. The Eleventh International Conference on Learning Representations , year=

  32. [32]

    Advances in neural information processing systems , volume=

    Language models are few-shot learners , author=. Advances in neural information processing systems , volume=

  33. [33]

    GPT-4 Technical Report

    Gpt-4 technical report. arxiv 2303.08774 , author=. View in Article , volume=

    work page internal anchor Pith review Pith/arXiv arXiv

  34. [34]

    Emergent Abilities of Large Language Models

    Emergent abilities of large language models , author=. arXiv preprint arXiv:2206.07682 , year=

    work page internal anchor Pith review Pith/arXiv arXiv

  35. [35]

    arXiv preprint arXiv:2310.05063 , year=

    Pushing the Limits of Pre-training for Time Series Forecasting in the CloudOps Domain , author=. arXiv preprint arXiv:2310.05063 , year=

    work page arXiv

  36. [36]

    On the Opportunities and Risks of Foundation Models

    On the opportunities and risks of foundation models , author=. arXiv preprint arXiv:2108.07258 , year=

    work page internal anchor Pith review Pith/arXiv arXiv

  37. [37]

    Proceedings of the IEEE international conference on computer vision , pages=

    Aligning books and movies: Towards story-like visual explanations by watching movies and reading books , author=. Proceedings of the IEEE international conference on computer vision , pages=

  38. [38]

    LLaMA: Open and Efficient Foundation Language Models

    Llama: Open and efficient foundation language models , author=. arXiv preprint arXiv:2302.13971 , year=

    work page internal anchor Pith review Pith/arXiv arXiv

  39. [39]

    LoRA: Low-Rank Adaptation of Large Language Models

    Lora: Low-rank adaptation of large language models , author=. arXiv preprint arXiv:2106.09685 , year=

    work page internal anchor Pith review Pith/arXiv arXiv

  40. [40]

    A Very British Affair: Six Britons and the Development of Time Series Analysis During the 20th Century , pages=

    Box and Jenkins: time series analysis, forecasting and control , author=. A Very British Affair: Six Britons and the Development of Time Series Analysis During the 20th Century , pages=. 2013 , publisher=

    2013

  41. [41]

    2018 , publisher=

    Improving language understanding by generative pre-training , author=. 2018 , publisher=

    2018

  42. [42]

    OpenAI blog , volume=

    Language models are unsupervised multitask learners , author=. OpenAI blog , volume=

  43. [43]

    Proceedings of the AAAI conference on artificial intelligence , volume=

    Informer: Beyond efficient transformer for long sequence time-series forecasting , author=. Proceedings of the AAAI conference on artificial intelligence , volume=

  44. [44]

    International Conference on Machine Learning , pages=

    What language model architecture and pretraining objective works best for zero-shot generalization? , author=. International Conference on Machine Learning , pages=. 2022 , organization=

    2022

  45. [45]

    arXiv preprint arXiv:2212.10559 , year=

    Why can gpt learn in-context? language models secretly perform gradient descent as meta optimizers , author=. arXiv preprint arXiv:2212.10559 , year=

    work page arXiv

  46. [46]

    Management science , volume=

    Forecasting sales by exponentially weighted moving averages , author=. Management science , volume=. 1960 , publisher=

    1960

  47. [47]

    2015 , publisher=

    Time series analysis: forecasting and control , author=. 2015 , publisher=

    2015

  48. [48]

    Advances in neural information processing systems , volume=

    A neural probabilistic language model , author=. Advances in neural information processing systems , volume=

  49. [49]

    2012 , publisher=

    Time series analysis by state space methods , author=. 2012 , publisher=

    2012

  50. [50]

    Journal of Machine Learning Research , volume=

    Palm: Scaling language modeling with pathways , author=. Journal of Machine Learning Research , volume=

  51. [51]

    A Survey of Large Language Models

    A survey of large language models , author=. arXiv preprint arXiv:2303.18223 , year=

    work page internal anchor Pith review Pith/arXiv arXiv

  52. [52]

    Training Compute-Optimal Large Language Models

    Training compute-optimal large language models , author=. arXiv preprint arXiv:2203.15556 , year=

    work page internal anchor Pith review Pith/arXiv arXiv

  53. [53]

    Advances in Neural Information Processing Systems , volume=

    Flamingo: a visual language model for few-shot learning , author=. Advances in Neural Information Processing Systems , volume=

  54. [54]

    Visual Instruction Tuning

    Visual instruction tuning , author=. arXiv preprint arXiv:2304.08485 , year=

    work page internal anchor Pith review Pith/arXiv arXiv

  55. [55]

    An Image is Worth 16x16 Words: Transformers for Image Recognition at Scale

    An image is worth 16x16 words: Transformers for image recognition at scale , author=. arXiv preprint arXiv:2010.11929 , year=

    work page internal anchor Pith review Pith/arXiv arXiv 2010

  56. [56]

    International conference on machine learning , pages=

    Learning transferable visual models from natural language supervision , author=. International conference on machine learning , pages=. 2021 , organization=

    2021

  57. [57]

    International Conference on Machine Learning , pages=

    Blip: Bootstrapping language-image pre-training for unified vision-language understanding and generation , author=. International Conference on Machine Learning , pages=. 2022 , organization=

    2022

  58. [58]

    arXiv preprint arXiv:2304.08424 , year=

    Long-term Forecasting with TiDE: Time-series Dense Encoder , author=. arXiv preprint arXiv:2304.08424 , year=

    work page arXiv

  59. [59]

    Advances in Neural Information Processing Systems , volume=

    Non-stationary transformers: Exploring the stationarity in time series forecasting , author=. Advances in Neural Information Processing Systems , volume=

  60. [60]

    International Conference on Machine Learning , pages=

    Fedformer: Frequency enhanced decomposed transformer for long-term series forecasting , author=. International Conference on Machine Learning , pages=. 2022 , organization=

    2022

  61. [61]

    Reformer: The Efficient Transformer

    Reformer: The efficient transformer , author=. arXiv preprint arXiv:2001.04451 , year=

    work page internal anchor Pith review Pith/arXiv arXiv 2001

  62. [62]

    The Eleventh International Conference on Learning Representations , year=

    Micn: Multi-scale local and global context modeling for long-term series forecasting , author=. The Eleventh International Conference on Learning Representations , year=

  63. [63]

    Proceedings of the 2000 ACM SIGMOD international conference on Management of data , pages=

    LOF: identifying density-based local outliers , author=. Proceedings of the 2000 ACM SIGMOD international conference on Management of data , pages=

    2000

  64. [64]

    2017 International joint conference on neural networks (IJCNN) , pages=

    Time series classification from scratch with deep neural networks: A strong baseline , author=. 2017 International joint conference on neural networks (IJCNN) , pages=. 2017 , organization=

    2017

  65. [65]

    BERT: Pre-training of Deep Bidirectional Transformers for Language Understanding

    Bert: Pre-training of deep bidirectional transformers for language understanding , author=. arXiv preprint arXiv:1810.04805 , year=

    work page internal anchor Pith review Pith/arXiv arXiv

  66. [66]

    Proceedings of the IEEE/CVF conference on computer vision and pattern recognition , pages=

    Do better imagenet models transfer better? , author=. Proceedings of the IEEE/CVF conference on computer vision and pattern recognition , pages=

  67. [67]

    arXiv preprint arXiv:2302.00861 , year=

    SimMTM: A Simple Pre-Training Framework for Masked Time-Series Modeling , author=. arXiv preprint arXiv:2302.00861 , year=

    work page arXiv

  68. [68]

    arXiv preprint arXiv:2202.01575 , year=

    CoST: Contrastive learning of disentangled seasonal-trend representations for time series forecasting , author=. arXiv preprint arXiv:2202.01575 , year=

    work page arXiv

  69. [69]

    Advances in neural information processing systems , volume=

    Attention is all you need , author=. Advances in neural information processing systems , volume=

  70. [70]

    A decoder-only foundation model for time-series forecasting

    A decoder-only foundation model for time-series forecasting , author=. arXiv preprint arXiv:2310.10688 , year=

    work page internal anchor Pith review Pith/arXiv arXiv

  71. [71]

    NIPS 2005 workshop on transfer learning , volume=

    To transfer or not to transfer , author=. NIPS 2005 workshop on transfer learning , volume=

    2005

  72. [72]

    BEiT: BERT Pre-Training of Image Transformers

    Beit: Bert pre-training of image transformers , author=. arXiv preprint arXiv:2106.08254 , year=

    work page internal anchor Pith review Pith/arXiv arXiv

  73. [73]

    VideoGPT: Video Generation using VQ-VAE and Transformers

    Videogpt: Video generation using vq-vae and transformers , author=. arXiv preprint arXiv:2104.10157 , year=

    work page internal anchor Pith review Pith/arXiv arXiv

  74. [74]

    Proceedings of the IEEE/CVF conference on computer vision and pattern recognition , pages=

    Momentum contrast for unsupervised visual representation learning , author=. Proceedings of the IEEE/CVF conference on computer vision and pattern recognition , pages=

  75. [75]

    International conference on machine learning , pages=

    A simple framework for contrastive learning of visual representations , author=. International conference on machine learning , pages=. 2020 , organization=

    2020

  76. [76]

    Proceedings of the 27th ACM SIGKDD conference on knowledge discovery & data mining , pages=

    A transformer-based framework for multivariate time series representation learning , author=. Proceedings of the 27th ACM SIGKDD conference on knowledge discovery & data mining , pages=

  77. [77]

    Advances in Neural Information Processing Systems , volume=

    Learning latent seasonal-trend representations for time series forecasting , author=. Advances in Neural Information Processing Systems , volume=

  78. [78]

    Proceedings of the AAAI Conference on Artificial Intelligence , volume=

    Ts2vec: Towards universal representation of time series , author=. Proceedings of the AAAI Conference on Artificial Intelligence , volume=

  79. [79]

    Advances in Neural Information Processing Systems , volume=

    Self-supervised contrastive pre-training for time series via time-frequency consistency , author=. Advances in Neural Information Processing Systems , volume=

  80. [80]

    arXiv preprint arXiv:2311.01933 , year=

    ForecastPFN: Synthetically-Trained Zero-Shot Forecasting , author=. arXiv preprint arXiv:2311.01933 , year=

    work page arXiv

Showing first 80 references.