Successfully completed simulations: {{ total }}. Buckets: {{ bucket_count }}. Every completed simulation belongs to one consecutive bucket in simulation order, oldest first. The final bucket includes any remaining runs.
Lines show mean Simulation Runtime and mean LLM Time per bucket, in minutes on the left axis. LLM Time includes prompt-to-response intervals and retries; runs without an LLM conversation contribute zero. Missing values are excluded from each statistic, with gaps when no values are available. Hover for simulation ranges, counts, mean and median high scores, and step summaries. Steps per Simulation shows the mean total steps per run on the right axis, using the same buckets. Mean steps per episode averages each run’s steps per episode equally. Reload to update.
No successfully completed simulations recorded yet.
{% endif %}Mean of each simulation's total steps divided by its runtime in seconds, with equal weight for each simulation in the bucket. Only runs with recorded steps and a positive runtime contribute; LLM time is excluded.
{% if rate_chart %} {{ rate_chart | safe }} {% else %}No successfully completed simulations recorded yet.
{% endif %}