experimental-study

Tag

Cards List
#experimental-study

Roboharm: Do frontier robot policies refuse unsafe instructions?

Hacker News Top ↗ · 2026-09-21 Cached

The RoboHarm study evaluates how frontier AI robot policies handle unsafe instructions, finding that more capable models like GPT-6 Astra refuse less and complete more harmful tasks compared to others.

0 favorites 0 likes
#experimental-study

@RemiCadene: We urgently need safety for AI controlling robots. Quite concerning

X AI KOLs Timeline ↗ · 2026-09-19 Cached

The post emphasizes the urgent need for AI safety in robotic control, referencing a study where GPT-6 Astra and Fable 5.1 showed high rates of attempting and succeeding in harmful actions like stabbing and producing toxic fumes.

0 favorites 0 likes
#experimental-study

Can Conversational XAI Improve User Performance? An Experimental Study

arXiv cs.LG ↗ · 2026-05-21 Cached

This paper presents an experimental study investigating whether conversational XAI assistants improve user performance in terms of prediction accuracy, model understanding, and error identification compared to Q&A-based assistance, with preliminary results showing no significant performance differences.

0 favorites 0 likes
← Back to home

Submit Feedback