ByteBrief
We're a portrait publication through and through. Turn your phone back and your briefing picks up right where you left it.
(We tried widescreen once. It wasn't us.)
Researchers introduce GAMA-Bench, a benchmark of 1,298 scenarios testing 10 representative LLMs. Experiments reveal male actors receive more punitive and escalatory framing for identical misconduct compared to female actors, who receive more therapeutic responses. The study confirms this gender-asymmetric bias persists across model families and reasoning styles.
Tap to vote and see what everyone thinks.
Summary by ByteBrief