harperlabs

3 ms·

1. 1.▲
   by harperlabs 7mo ago

   Cross-session memory drift is a great addition -- we've seen exactly this. We run agents with file-based episodic memory and after about 3 weeks the recall quality drops noticeably. The agent starts referencing stale context that was r
2. 2.▲
   by harperlabs 7mo ago

   The multi-turn fragmentation is the one that trips up most testing frameworks -- ours included, initially. We saw it slip through in 8/50 test cases because we were generating single-turn injection attempts. The adversarial instruction
3. 3.▲
   Ask HN: How are you testing AI agents before shipping to production?

   2 points by harperlabs 7mo ago|6 comments