SimCoachCorpus: Naturalistic Language and Trajectories for Embodied Teaching

SimCoachCorpus is a high-quality embodied interactive teaching dataset in the high-performance driving domain. It was authored by Emily Sumner and published at the 32nd ACM SIGKDD Conference on Knowledge Discovery and Data Mining in 2026. The dataset was last updated on June 1, 2026.

Use Cases

• Train models for natural language instruction in embodied teaching based on the described interactive teaching data.
• Analyze trajectories for agent behavior in high-performance driving simulations based on the dataset's domain.
• Develop multimodal AI systems combining language and motion based on the dataset's likely content.
• Benchmark interactive teaching agents in simulated environments based on the dataset's described purpose.

Strengths

• Dataset is described as 'high-quality' in the input description.
• Published at a major conference (ACM SIGKDD 2026), suggesting peer review.
• Last update timestamp is precisely recorded as 2026-06-01 15:04:58.

Limitations

• Column-level documentation is absent; field semantics must be inferred after download.
• Row count is unknown, which may limit suitability assessment.
• File formats and data size are unspecified, complicating download planning.

Provenance

Source

Sim Coach Dataverse

Collection Method

Likely collected from an embodied interactive teaching simulation.

Freshness

Last updated 2026-06-01 15:04:58.