Is One Layer Enough? Understanding Inference Dynamics in Tabular Foundation Models

• URL: http://arxiv.org/abs/2605.06510v1
• Date: Thu, 07 May 2026 16:22:04 GMT
• Title: Is One Layer Enough? Understanding Inference Dynamics in Tabular Foundation Models
• Abstract summary: We present the first large-scale mechanistic study of layerwise dynamics in 6 state-of-the-art tabular in-context learning models.<n>Our findings indicate substantial depthwise redundancy across multiple models, suggesting iterative refinement with overlapping computations during inference stages.<n>We design a proof-of-concept, looped single-layer model that uses only 20% of the original model's parameters while achieving comparable performance.
• Score: 2.8473361448579713
• License: http://arxiv.org/licenses/nonexclusive-distrib/1.0/
• Abstract: Transformer-based tabular foundation models (TFMs) dominate small to medium tabular predictive benchmark tasks, yet their inference mechanisms remain largely unexplored. We present the first large-scale mechanistic study of layerwise dynamics in 6 state-of-the-art tabular in-context learning models. We explore how predictions emerge across depth, identify distinct stages of inference and reveal latent-space dynamics that differ from those of language models. Our findings indicate substantial depthwise redundancy across multiple models, suggesting iterative refinement with overlapping computations during inference stages. Guided by these insights, we design a proof-of-concept, looped single-layer model that uses only 20% of the original model's parameters while achieving comparable performance. The code is available at https://github.com/amirbalef/is_one_layer_enough.

Related papers

• TabPFN Through The Looking Glass: An interpretability study of TabPFN and its internal representations [1.3986052226424095]
  We analyze the information encoded inside the model's hidden representations.<n>We observe clear signals that correspond to both intermediate and final quantities involved in the model's prediction process.
  arXiv  Detail & Related papers  (2026-01-13T03:14:25Z)
• Large EEG-U-Transformer for Time-Step Level Detection Without Pre-Training [1.3254304182988286]
  We propose a simple U-shaped model to efficiently learn representations by capturing both local and global features.<n>Compared to other window-level classification models, our method directly outputs predictions at the time-step level.<n>Our model won 1st place in the 2025 "seizure detection challenge" organized in the International Conference on Artificial Intelligence in Epilepsy and Other Neurological Disorders.
  arXiv  Detail & Related papers  (2025-04-01T01:33:42Z)
• LESA: Learnable LLM Layer Scaling-Up [57.0510934286449]
  Training Large Language Models (LLMs) from scratch requires immense computational resources, making it prohibitively expensive.<n>Model scaling-up offers a promising solution by leveraging the parameters of smaller models to create larger ones.<n>We propose textbfLESA, a novel learnable method for depth scaling-up.
  arXiv  Detail & Related papers  (2025-02-19T14:58:48Z)
• Rethinking Weight-Averaged Model-merging [15.2881959315021]
  Model merging, particularly through weight averaging, has shown surprising effectiveness in saving computations and improving model performance without any additional training.<n>In this work, we reinterpret weight-averaged model merging through the lens of interpretability and provide empirical insights into the underlying mechanisms that govern its behavior.
  arXiv  Detail & Related papers  (2024-11-14T08:02:14Z)
• Revisiting Nearest Neighbor for Tabular Data: A Deep Tabular Baseline Two Decades Later [76.66498833720411]
  We introduce a differentiable version of $K$-nearest neighbors (KNN) originally designed to learn a linear projection to capture semantic similarities between instances.<n>Surprisingly, our implementation of NCA using SGD and without dimensionality reduction already achieves decent performance on tabular data.<n>We conclude our paper by analyzing the factors behind these improvements, including loss functions, prediction strategies, and deep architectures.
  arXiv  Detail & Related papers  (2024-07-03T16:38:57Z)
• Causal Estimation of Memorisation Profiles [58.20086589761273]
  Understanding memorisation in language models has practical and societal implications. Memorisation is the causal effect of training with an instance on the model's ability to predict that instance. This paper proposes a new, principled, and efficient method to estimate memorisation based on the difference-in-differences design from econometrics.
  arXiv  Detail & Related papers  (2024-06-06T17:59:09Z)
• LaCo: Large Language Model Pruning via Layer Collapse [56.92068213969036]
  Large language models (LLMs) based on transformer are witnessing a notable trend of size expansion. Existing methods such as model quantization, knowledge distillation, and model pruning are constrained by various issues. We propose a concise layer-wise structured pruner called textitLayer Collapse (LaCo), in which rear model layers collapse into a prior layer.
  arXiv  Detail & Related papers  (2024-02-17T04:16:30Z)
• Layer-wise Linear Mode Connectivity [52.6945036534469]
  Averaging neural network parameters is an intuitive method for the knowledge of two independent models. It is most prominently used in federated learning. We analyse the performance of the models that result from averaging single, or groups.
  arXiv  Detail & Related papers  (2023-07-13T09:39:10Z)
• Bayesian Beta-Bernoulli Process Sparse Coding with Deep Neural Networks [11.937283219047984]
  Several approximate inference methods have been proposed for deep discrete latent variable models. We propose a non-parametric iterative algorithm for learning discrete latent representations in such deep models. We evaluate our method across datasets with varying characteristics and compare our results to current amortized approximate inference methods.
  arXiv  Detail & Related papers  (2023-03-14T20:50:12Z)
• Training Trajectories of Language Models Across Scales [99.38721327771208]
  Scaling up language models has led to unprecedented performance gains. How do language models of different sizes learn during pre-training? Why do larger language models demonstrate more desirable behaviors?
  arXiv  Detail & Related papers  (2022-12-19T19:16:29Z)
• Dynamic Latent Separation for Deep Learning [67.62190501599176]
  A core problem in machine learning is to learn expressive latent variables for model prediction on complex data. Here, we develop an approach that improves expressiveness, provides partial interpretation, and is not restricted to specific applications.
  arXiv  Detail & Related papers  (2022-10-07T17:56:53Z)
• Consistent Counterfactuals for Deep Models [25.1271020453651]
  Counterfactual examples are used to explain predictions of machine learning models in key areas such as finance and medical diagnosis. This paper studies the consistency of model prediction on counterfactual examples in deep networks under small changes to initial training conditions.
  arXiv  Detail & Related papers  (2021-10-06T23:48:55Z)

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.