Reproduces the full prefill sensitivity analysis pipeline for reward hacking indicators. Use when evaluating how suscep…
Reproduces the full prefill sensitivity analysis pipeline for reward hacking indicators. Use when evaluating how susceptible model checkpoints are to exploit-eliciting prefills, computing token-based trajectories, or comparing logprob vs token-count as predictors of exploitability.
aiskillstore
cli
free
Others in the same category, ranked by how often they are opened.
