The Few Govern the Many:Unveiling Few-Layer Dominance for Time Series Models

• URL: http://arxiv.org/abs/2511.07237v1
• Date: Mon, 10 Nov 2025 15:53:29 GMT
• Title: The Few Govern the Many:Unveiling Few-Layer Dominance for Time Series Models
• Authors: Xin Qiu, Junlong Tong, Yirong Sun, Yunpu Ma, Xiaoyu Shen,
• Abstract summary: We observe a puzzling phenomenon that larger models do emphNOT achieve better performance.<n>We propose a practical method to automatically identify and retain only these dominant layers.
• Score: 14.479996384788445
• License: http://arxiv.org/licenses/nonexclusive-distrib/1.0/
• Abstract: Large-scale models are at the forefront of time series (TS) forecasting, dominated by two paradigms: fine-tuning text-based Large Language Models (LLM4TS) and training Time Series Foundation Models (TSFMs) from scratch. Both approaches share a foundational assumption that scaling up model capacity and data volume leads to improved performance. However, we observe a \textit{\textbf{scaling paradox}} in TS models, revealing a puzzling phenomenon that larger models do \emph{NOT} achieve better performance. Through extensive experiments on two model families across four scales (100M to 1.7B parameters) and diverse data (up to 6B observations), we rigorously confirm that the scaling paradox is a pervasive issue. We then diagnose its root cause by analyzing internal representations, identifying a phenomenon we call \textit{few-layer dominance}: only a small subset of layers are functionally important, while the majority are redundant, under-utilized, and can even distract training. Based on this discovery, we propose a practical method to automatically identify and retain only these dominant layers. In our models, retaining only 21\% of the parameters achieves up to a 12\% accuracy improvement and a 2.7$\times$ inference speedup. We validate the universality of our method on 8 prominent SOTA models (LLM4TS and TSFMs, 90M to 6B), showing that retaining less than 30\% of layers achieves comparable or superior accuracy in over 95\% of tasks.

Related papers

• LimiX: Unleashing Structured-Data Modeling Capability for Generalist Intelligence [61.46575527504109]
  LimiX-16M and LimiX-2M treat structured data as a joint distribution over variables and missingness.<n>We evaluate LimiX models across 11 large structured-data benchmarks with broad regimes of sample size, feature dimensionality, class number, categorical-to-numerical feature ratio, missingness, and sample-to-feature ratios.
  arXiv  Detail & Related papers  (2025-09-03T17:39:08Z)
• Compute-Optimal Scaling for Value-Based Deep RL [99.680827753493]
  We investigate compute scaling for online, value-based deep RL.<n>Our analysis reveals a nuanced interplay between model size, batch size, and UTD.<n>We provide a mental model for understanding this phenomenon and build guidelines for choosing batch size and UTD.
  arXiv  Detail & Related papers  (2025-08-20T17:54:21Z)
• Intention-Conditioned Flow Occupancy Models [80.42634994902858]
  Large-scale pre-training has fundamentally changed how machine learning research is done today.<n>Applying this same framework to reinforcement learning is appealing because it offers compelling avenues for addressing core challenges in RL.<n>Recent advances in generative AI have provided new tools for modeling highly complex distributions.
  arXiv  Detail & Related papers  (2025-06-10T15:27:46Z)
• Output Scaling: YingLong-Delayed Chain of Thought in a Large Pretrained Time Series Forecasting Model [55.25659103706409]
  This framework achieves state-of-the-art performance for our designed foundation model, YingLong.<n>YingLong is a non-causal, bidirectional attention encoder-only transformer trained through masked token recovery.<n>We release four foundation models ranging from 6M to 300M parameters, demonstrating superior results in zero-shot tasks.
  arXiv  Detail & Related papers  (2025-05-20T14:31:06Z)
• TSPulse: Dual Space Tiny Pre-Trained Models for Rapid Time-Series Analysis [12.034816114258803]
  TSPulse is an ultra-compact time-series pre-trained model with only 1M parameters.<n>It performs strongly across classification, anomaly detection, imputation, and retrieval tasks.<n>Results are achieved with just 1M parameters (10-100X smaller than existing SOTA models)
  arXiv  Detail & Related papers  (2025-05-19T12:18:53Z)
• Scaling Inference-Efficient Language Models [3.271571137474847]
  We show that model architecture affects inference latency, where models of the same size can have up to 3.5x difference in latency.<n>We modify the Chinchilla scaling laws to co-optimize the model parameter count, the number of training tokens, and the model architecture.<n>We release the Morph-1B model, which improves inference latency by 1.8x while maintaining accuracy on downstream tasks.
  arXiv  Detail & Related papers  (2025-01-30T03:16:44Z)
• Truncated Consistency Models [57.50243901368328]
  Training consistency models requires learning to map all intermediate points along PF ODE trajectories to their corresponding endpoints.<n>We empirically find that this training paradigm limits the one-step generation performance of consistency models.<n>We propose a new parameterization of the consistency function and a two-stage training procedure that prevents the truncated-time training from collapsing to a trivial solution.
  arXiv  Detail & Related papers  (2024-10-18T22:38:08Z)
• Memory-Efficient Fine-Tuning for Quantized Diffusion Model [12.875837358532422]
  We introduce TuneQDM, a memory-efficient fine-tuning method for quantized diffusion models. Our method consistently outperforms the baseline in both single-/multi-subject generations.
  arXiv  Detail & Related papers  (2024-01-09T03:42:08Z)
• Tiny Time Mixers (TTMs): Fast Pre-trained Models for Enhanced Zero/Few-Shot Forecasting of Multivariate Time Series [11.635608108358575]
  We introduce Tiny Time Mixers (TTM), a compact model with effective transfer learning capabilities, trained exclusively on public TS datasets. TTM incorporates innovations like adaptive patching, diverse resolution sampling, and resolution prefix tuning to handle pre-training on varied dataset resolutions. It outperforms existing popular benchmarks in zero/few-shot forecasting by (4-40%), while reducing computational requirements significantly.
  arXiv  Detail & Related papers  (2024-01-08T15:21:21Z)
• Model soups: averaging weights of multiple fine-tuned models improves accuracy without increasing inference time [69.7693300927423]
  We show that averaging the weights of multiple models fine-tuned with different hyper parameter configurations improves accuracy and robustness. We show that the model soup approach extends to multiple image classification and natural language processing tasks.
  arXiv  Detail & Related papers  (2022-03-10T17:03:49Z)
• No One Representation to Rule Them All: Overlapping Features of Training Methods [12.58238785151714]
  High-performing models tend to make similar predictions regardless of training methodology. Recent work has made very different training techniques, such as large-scale contrastive learning, yield competitively-high accuracy. We show these models specialize in generalization of the data, leading to higher ensemble performance.
  arXiv  Detail & Related papers  (2021-10-20T21:29:49Z)
• Model-based micro-data reinforcement learning: what are the crucial model properties and which model to choose? [0.2836066255205732]
  We contribute to micro-data model-based reinforcement learning (MBRL) by rigorously comparing popular generative models. We find that on an environment that requires multimodal posterior predictives, mixture density nets outperform all other models by a large margin. We also found that deterministic models are on par, in fact they consistently (although non-significantly) outperform their probabilistic counterparts.
  arXiv  Detail & Related papers  (2021-07-24T11:38:25Z)
• Exploring Sparse Expert Models and Beyond [51.90860155810848]
  Mixture-of-Experts (MoE) models can achieve promising results with outrageous large amount of parameters but constant computation cost. We propose a simple method called expert prototyping that splits experts into different prototypes and applies $k$ top-$1$ routing. This strategy improves the model quality but maintains constant computational costs, and our further exploration on extremely large-scale models reflects that it is more effective in training larger models.
  arXiv  Detail & Related papers  (2021-05-31T16:12:44Z)

This list is automatically generated from the titles and abstracts of the papers in this site.

This site does not guarantee the quality of this site (including all information) and is not responsible for any consequences.