Davide Paglieri

@paglieridavide

Research Scientist @GoogleDeepMind Previously: @UCL_DARK, @bendingspoons

following
followers

A proposal to help AI agent swarms police themselves

A DeepMind Institute essay coauthor says honest agents try to flag cheating peers when given transparent channels—a behavior the proposed self-policing tools would support.

6 Sources, 26d ago, updated 26d ago

Claude Opus 5 and GPT 5.6 Terra join the BALROG leaderboard

A post announcing the update places Terra between Luna and Sol and says Opus 5 comes close to GPT-6 astra, but falls short.

2 Sources, 30d ago, first seen 30d ago

Featured in

A proposal to help AI agent swarms police themselves

Claude Opus 5 and GPT 5.6 Terra join the BALROG leaderboard

Featured in

A proposal to help AI agent swarms police themselves

Claude Opus 5 and GPT 5.6 Terra join the BALROG leaderboard