Gemma The Unstopping: a Behavioral Experiment

·LessWrong··

Following up on my previous experiment - studying Gemma's behavior on agentic tasks when given the number of steps left across the run - I endeavored to see whether giving Gemma a stop_run tool meaningfully changes it's behavior.My initial assumption was that it would not change the completion rate of these tasks, and that has been correct.What I wanted to test:Whether Gemma would "give up" when nearing the end of the run after hitting a dead end. Interesting.Whether Gemma would "want to stop" -...

Read full article →

Related Articles

EPA says power for data centers can sidestep pollution laws
Levitating · Hacker News · 5h ago
GLM-5.3 is now open-weight
jeudesprits · Hacker News · 3h ago
Pentagon's blacklisting of Anthropic was unlawful, US judge rules
softwaredoug · Hacker News · 7h ago
Autistici/Inventati's main .org domain goes dark after US terrorism designation
subvol · Hacker News · 4h ago
Saving 100 terabytes of memory by optimizing 1.1.1.1's DNS cache
TangerineDream · Hacker News · 1d ago