ariaw

• ariaw 11 Jan 2026 21:17 UTC

  1 point

  0

  in reply to: Shashwat Saxena’s comment on: Steer­ing RL Train­ing: Bench­mark­ing In­ter­ven­tions Against Re­ward Hacking

  It’s hard for me to help without more information. I’ve responded to your email asking to send some of the files created by training, I can try to help you debug from there.

• ariaw 8 Jan 2026 0:39 UTC

  3 points

  0

  in reply to: Shashwat Saxena’s comment on: Steer­ing RL Train­ing: Bench­mark­ing In­ter­ven­tions Against Re­ward Hacking

  Sorry about that! The repository was updated a few days ago to fix this. Let me know if you have any further issues!

• ariaw 5 Jan 2026 17:15 UTC

  3 points

  0

  in reply to: Jozdien’s comment on: Steer­ing RL Train­ing: Bench­mark­ing In­ter­ven­tions Against Re­ward Hacking

  This chart is definitely related to understanding test-time hacking for one inoculation prompt, but this figure shows the reward hacking rates when the model is specifically instructed not to hack. I would assume the model hacks quite a bit more without the no-hack prompt, but I don’t see those results directly stated in the paper (correct me if I am wrong).
  
  Our results on inoculation prompting at test-time are shown with the inoculation prompt removed, but we do not add a no-hack system prompt. To be clear, we see our results on inoculation prompting as in-line with Anthropic’s findings.

Steer­ing RL Train­ing: Bench­mark­ing In­ter­ven­tions Against Re­ward Hacking