Geographic Diversity Beats Data Volume for Cross-Domain Generalization in Zero-Label JEPA Driving World Models

• URL: http://arxiv.org/abs/2607.04500v1
• Date: Sun, 05 Jul 2026 20:58:16 GMT
• Title: Geographic Diversity Beats Data Volume for Cross-Domain Generalization in Zero-Label JEPA Driving World Models
• Abstract summary: Self-supervised latent world models can assign a surprise score to driving scenarios without any human labels.<n>We study whether such a model, trained on driving data from one geographic region, can generalize its notion of complexity to unseen cities and sensor configurations.<n>We find that models trained on geographically diverse data generalize significantly better than models trained on equal amounts of single-geography data.
• Score: 0.0
• License: http://creativecommons.org/licenses/by/4.0/
• Abstract: Self-supervised latent world models can assign a surprise score to driving scenarios without any human labels. A natural follow-up question is whether such a model, trained on driving data from one geographic region, can generalize its notion of complexity to unseen cities and sensor configurations. We study this question through a controlled transfer experiment: we train JEPA-based world models on nuPlan data (Pittsburgh, Boston, Singapore) and evaluate zero-shot on held-out Argoverse 2 validation scenarios from Miami and Austin. We find that models trained on geographically diverse data generalize significantly better than models trained on equal amounts of single-geography data. In a matched-scale ablation at 63,000 scenarios per condition (n=3 seeds each), combined training reduces mean surprise score by 16.5% relative to nuPlan-only training (0.228 +/- 0.015 vs 0.273 +/- 0.008). Notably, training on 200,000 AV2-only scenarios (3x more data from one geography) still produces higher surprise (0.264) than the combined 63K model, suggesting that geographic diversity is a stronger predictor of cross-domain generalization than raw data volume.

Related papers

• Zero-Label Driving Scenario Complexity Detection via Joint Embedding Predictive Architecture [0.0]
  We ask whether a model can discover scenario complexity on its own, with no labels at any stage.<n>We train a minimal Joint Embedding Predictive Architecture (JEPA) on structured agent state data from the nuPlan mini dataset.<n>Without access to any ground-truth labels during training or evaluation setup, the model assigns significantly higher scores to scenarios involving unprotected turns, crosswalk interactions, and pedestrian proximity.
  arXiv  Detail & Related papers  (2026-06-21T05:26:58Z)
• Zero-Shot Cross-City Generalization in End-to-End Autonomous Driving: Self-Supervised versus Supervised Representations [9.18632648031395]
  We investigate zero-shot cross-city generalization in end-to-end trajectory planning.<n>Self-supervised visual representations improve transfer across cities.<n>These results establish zero-shot geographic transfer as a necessary test for evaluating end-to-end autonomous driving systems.
  arXiv  Detail & Related papers  (2026-03-12T01:19:32Z)
• Metadata Conditioned Large Language Models for Localization [25.913929585741034]
  We show that metadata conditioning consistently improves in-region performance without sacrificing cross-region generalization.<n>Our ablation studies demonstrate that URL-level metadata alone captures much of the geographic signal.<n>After instruction tuning, metadata conditioned global models achieve accuracy comparable to LLaMA-3.2-1B-Instruct, despite being trained on substantially less data.
  arXiv  Detail & Related papers  (2026-01-21T18:20:59Z)
• Prismatic Synthesis: Gradient-based Data Diversification Boosts Generalization in LLM Reasoning [77.120955854093]
  We show that data diversity can be a strong predictor of generalization in language models.<n>We introduce G-Vendi, a metric that quantifies diversity via the entropy of model-induced gradients.<n>We present Prismatic Synthesis, a framework for generating diverse synthetic data.
  arXiv  Detail & Related papers  (2025-05-26T16:05:10Z)
• WorldPM: Scaling Human Preference Modeling [130.23230492612214]
  We propose World Preference Modeling$ (WorldPM) to emphasize this scaling potential.<n>We collect preference data from public forums covering diverse user communities.<n>We conduct extensive training using 15M-scale data across models ranging from 1.5B to 72B parameters.
  arXiv  Detail & Related papers  (2025-05-15T17:38:37Z)
• No "Zero-Shot" Without Exponential Data: Pretraining Concept Frequency Determines Multimodal Model Performance [68.18779562801762]
  multimodal models require exponentially more data to achieve linear improvements in downstream "zero-shot" performance. Our study reveals an exponential need for training data which implies that the key to "zero-shot" generalization capabilities under large-scale training paradigms remains to be found.
  arXiv  Detail & Related papers  (2024-04-04T17:58:02Z)
• Evaluating the Role of Training Data Origin for Country-Scale Cropland Mapping in Data-Scarce Regions: A Case Study of Nigeria [0.6249768559720122]
  A key challenge is understanding how the quantity, quality, and proximity of the training data to the target region influences model performance.<n>We evaluate this in Nigeria, using 1,827 manually labelled samples covering the whole country, and subsets of the Geowiki dataset.<n>Results show local data significantly boosts performance, with accuracy gains up to 0.246 (RF) and 0.178 (LSTM)
  arXiv  Detail & Related papers  (2023-12-18T01:23:22Z)
• Strict baselines for Covid-19 forecasting and ML perspective for USA and Russia [105.54048699217668]
  Covid-19 allows researchers to gather datasets accumulated over 2 years and to use them in predictive analysis. We present the results of a consistent comparative study of different types of methods for predicting the dynamics of the spread of Covid-19 based on regional data for two countries: the United States and Russia.
  arXiv  Detail & Related papers  (2022-07-15T18:21:36Z)
• Dataset Cartography: Mapping and Diagnosing Datasets with Training Dynamics [118.75207687144817]
  We introduce Data Maps, a model-based tool to characterize and diagnose datasets. We leverage a largely ignored source of information: the behavior of the model on individual instances during training. Our results indicate that a shift in focus from quantity to quality of data could lead to robust models and improved out-of-distribution generalization.
  arXiv  Detail & Related papers  (2020-09-22T20:19:41Z)
• Think Locally, Act Globally: Federated Learning with Local and Global Representations [92.68484710504666]
  Federated learning is a method of training models on private data distributed over multiple devices. We propose a new federated learning algorithm that jointly learns compact local representations on each device. We also evaluate on the task of personalized mood prediction from real-world mobile data where privacy is key.
  arXiv  Detail & Related papers  (2020-01-06T12:40:21Z)

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.