rsi-index/post-training

ALE RSI post-training benchmark tasks for evaluating agent capabilities in research-level post-training workflows.

← Back to Registry

Run this task

CLI:

inspect eval inspect_harbor/rsi_index_post_training --model openai/gpt-5

Python:

from inspect_ai import eval
from inspect_harbor import rsi_index_post_training

eval(rsi_index_post_training(), model="openai/gpt-5")

Dataset information

Harbor registry	rsi-index/post-training
Inspect task	rsi_index_post_training
Latest digest	sha256:46e29b6a00e487dbf4ab88c40c1a525941df5f3d7ce6956b42c35d5432031bfb
Samples	1

See Task Parameters for the parameter set shared across all Harbor tasks.