Geometric Entropy: When Trajectory Diversity Helps and Hurts in Imitation Learning

• URL: http://arxiv.org/abs/2606.20871v1
• Date: Thu, 18 Jun 2026 19:02:08 GMT
• Title: Geometric Entropy: When Trajectory Diversity Helps and Hurts in Imitation Learning
• Abstract summary: We study how trajectory-shape diversity in demonstrations affects imitation learning (IL) performance across models, tasks, and data scales.<n>We introduce Geometric Entropy (H_G), a task-agnostic metric that quantifies the intrinsic diversity of transit trajectories.
• Score: 16.153147572760883
• License: http://arxiv.org/licenses/nonexclusive-distrib/1.0/
• Abstract: We study how trajectory-shape diversity in demonstrations affects imitation learning (IL) performance across models, tasks, and data scales. We introduce Geometric Entropy (H_G), a task-agnostic metric that quantifies the intrinsic diversity of transit trajectories after normalizing away extrinsic variation, such as goal pose and workspace scale, via target-frame alignment. Across multiple IL architectures and both simulated and real-robot contact-rich manipulation tasks, we observe a consistent inverted-U relationship between success and H_G: increasing geometric diversity improves robustness in low-diversity regimes but degrades performance once diversity induces strategy ambiguity. Moreover, the optimal entropy shifts toward lower values as task mastery increases through more data, easier tasks, or stronger priors, and for a pretrained vision-language-action model the trend becomes effectively monotonic decreasing. Practically, H_G enables fast pre-training auditing of demonstration datasets and offers a simple guideline for calibrating demonstrations toward the learnable regime.

Related papers

• SPARD: Self-Paced Curriculum for RL Alignment via Integrating Reward Dynamics and Data Utility [71.76390626651254]
  We propose SPARD, a framework that establishes an automated, self-paced curriculum by perceiving learning progress to adjust dynamically multi-objective reward weights and data importance.<n>Extensive experiments across multiple benchmarks demonstrate that SPARD significantly enhances model capabilities across all domains.
  arXiv  Detail & Related papers  (2026-04-09T05:37:22Z)
• One-Policy-Fits-All: Geometry-Aware Action Latents for Cross-Embodiment Manipulation [51.66470249744105]
  Cross-embodiment manipulation is crucial for enhancing the scalability of robot manipulation.<n>We propose OneFits-All (OPFA), a framework that enables learning a single, versatile policy across multiple embodiments.
  arXiv  Detail & Related papers  (2026-03-15T17:59:08Z)
• Diversity You Can Actually Measure: A Fast, Model-Free Diversity Metric for Robotics Datasets [12.45101201787042]
  We study how dataset diversity affects generalization performance in robot imitation learning.<n>We introduce FAKTUAL, a data curation algorithm that selects a subset of demonstrations maximizing entropy.<n>Our results suggest that FAKTUAL is a practical tool for understanding and improving dataset diversity in robot imitation learning.
  arXiv  Detail & Related papers  (2026-03-12T07:54:43Z)
• TopoCurate:Modeling Interaction Topology for Tool-Use Agent Training [53.93696896939915]
  Training tool-use agents typically rely on Supervised Fine-Tuning (SFT) on successful trajectories and Reinforcement Learning (RL) on pass-rate-selected tasks.<n>We propose TopoCurate, an interaction-aware framework that projects multi-trial rollouts from the same task into a unified semantic quotient topology.<n>TopoCurate achieves consistent gains of 4.2% (SFT) and 6.9% (RL) over state-of-the-art baselines.
  arXiv  Detail & Related papers  (2026-03-02T10:38:54Z)
• HGMP:Heterogeneous Graph Multi-Task Prompt Learning [18.703129208282913]
  We propose a novel multi-task prompt framework for the heterogeneous graph domain, named HGMP.<n>First, to bridge the gap between the pre-trained model and downstream tasks, we reformulate all downstream tasks into a unified graph-level task format.<n>We design a graph-level contrastive pre-training strategy to better leverage heterogeneous information and enhance performance in multi-task scenarios.
  arXiv  Detail & Related papers  (2025-07-10T04:01:47Z)
• Gradient Similarity Surgery in Multi-Task Deep Learning [1.2299544525529198]
  This work introduces a novel gradient surgery method based on a gradient magnitude similarity measure to guide the optimisation process.<n>The Similarity-Aware Momentum Gradient Surgery (SAM-GS) adopts gradient equalisation and modulation of the first-order momentum.
  arXiv  Detail & Related papers  (2025-06-06T14:40:50Z)
• Self-Controlled Dynamic Expansion Model for Continual Learning [10.447232167638816]
  This paper introduces an innovative Self-Controlled Dynamic Expansion Model (SCDEM)<n>SCDEM orchestrates multiple trainable pre-trained ViT backbones to furnish diverse and semantically enriched representations.<n>An extensive series of experiments have been conducted to evaluate the proposed methodology's efficacy.
  arXiv  Detail & Related papers  (2025-04-14T15:22:51Z)
• Beyond Losses Reweighting: Empowering Multi-Task Learning via the Generalization Perspective [61.10883077161432]
  Multi-task learning (MTL) trains deep neural networks to optimize several objectives simultaneously using a shared backbone.<n>We introduce a novel MTL framework that leverages weight perturbation to regulate gradient norms, thus improving generalization.<n>Our method significantly outperforms existing gradient-based MTL techniques in terms of task performance and overall model robustness.
  arXiv  Detail & Related papers  (2022-11-24T17:19:30Z)
• Revisiting Consistency Regularization for Semi-Supervised Learning [80.28461584135967]
  We propose an improved consistency regularization framework by a simple yet effective technique, FeatDistLoss. Experimental results show that our model defines a new state of the art for various datasets and settings.
  arXiv  Detail & Related papers  (2021-12-10T20:46:13Z)
• Visual Adversarial Imitation Learning using Variational Models [60.69745540036375]
  Reward function specification remains a major impediment for learning behaviors through deep reinforcement learning. Visual demonstrations of desired behaviors often presents an easier and more natural way to teach agents. We develop a variational model-based adversarial imitation learning algorithm.
  arXiv  Detail & Related papers  (2021-07-16T00:15:18Z)
• Task-Feature Collaborative Learning with Application to Personalized Attribute Prediction [166.87111665908333]
  We propose a novel multi-task learning method called Task-Feature Collaborative Learning (TFCL) Specifically, we first propose a base model with a heterogeneous block-diagonal structure regularizer to leverage the collaborative grouping of features and tasks. As a practical extension, we extend the base model by allowing overlapping features and differentiating the hard tasks.
  arXiv  Detail & Related papers  (2020-04-29T02:32:04Z)

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.