Lio@lzvck_Oct 6Shuga@shuga_vibesOct 6BUENO quien viene a jugar al CS 1.6, primera prueba abierta al público: sla-cs.vercel.app Entren a la sala SOLO HUMANOS pass: shugiendolosvibentimientos3
Shuga@shuga_vibesOct 6BUENO quien viene a jugar al CS 1.6, primera prueba abierta al público: sla-cs.vercel.app Entren a la sala SOLO HUMANOS pass: shugiendolosvibentimientos
Lio@lzvck_Sep 22Que baje el ritmo de training habian dicho?Claude@claudeaiSep 22Replying to @claudeaiOpus 5.5 is a major step up from Opus 5, leading on agentic coding, computer use, and knowledge work.11
Claude@claudeaiSep 22Replying to @claudeaiOpus 5.5 is a major step up from Opus 5, leading on agentic coding, computer use, and knowledge work.
Lio@lzvck_Sep 22We checked whether SI-agent papers already close the usual eval threats. Most don't measure it cleanly. Paper: seriora.ai/papers/self-im…Seriora Research@serioraresearchSep 22Most self-improving LLM-agent papers fail basic eval threats. We audited 41 peer-visible SI-agent papers (2024–2026) with a fixed 5-trap checklist.12
Seriora Research@serioraresearchSep 22Most self-improving LLM-agent papers fail basic eval threats. We audited 41 peer-visible SI-agent papers (2024–2026) with a fixed 5-trap checklist.
