Personalized Route Planning Benchmarks for LLM and Multi-Cost Simulation

PersonaRoute Bench provides datasets for training and testing personalized routing models, as presented in the 'PersonalizedRouter' paper. The repository, created by ulab-ai, includes data generated via multi-cost-efficiency and LLM-as-a-Judge simulation strategies. The dataset page was last updated in February 2026.

Use Cases

• Train a model to predict optimal routes using user-specific features from router_user_data_v1.
• Benchmark LLM-based routing agents against multi-cost simulation strategies using the v2 dataset variant.
• Evaluate model scalability on the 'large' dataset variant denoting a large-scale setting.
• Analyze the impact of different simulation strategies (v1 vs. v2) on route planning performance metrics.

Strengths

• Includes multiple dataset variants (v1, v2, large) for comparative analysis.
• Dataset page updated in February 2026, indicating recent maintenance.

Limitations

• Specific row counts, column names, and dataset size are unknown.
• Data scope and geographic coverage are not specified, limiting reproducibility assessment.

Provenance

Source

ulab-ai, associated with the 'PersonalizedRouter' research paper.

Collection Method

Data generated via simulation strategies (multi-cost-efficiency and LLM-as-a-Judge) as described in the paper.

Freshness

Last updated on the platform on 2026-02-22.