A user lists Grok-4.7 at 14/17 and calls this edition of the 'priv8 hacker jeopardy' benchmark saturated.
A user shared scores for what they call a saturated edition of the 'priv8 hacker jeopardy' benchmark: 14.5/17 for gemini-argon-hi and 14/17 for grok-4.7. They suggested it may be the last edition.
400
2 Sources, first seen ago
@tensecorrectionpriv8 hacker jeopardy benchmark, saturated edition: gemini-argon-hi 14.5/17 Sundar didn't raise no bitch grok-4.7 14/17 didn't even try 'til now, no surprise muse six-seven :( but it's a very good model I swear may be the last one: the attractors are very visible now
@zacharynadoRT @tensecorrection: priv8 hacker jeopardy benchmark, saturated edition: gemini-argon-hi 14.5/17 Sundar didn't raise no bitch grok-4…@tensecorrectionpriv8 hacker jeopardy benchmark, saturated edition: gemini-argon-hi 14.5/17 Sundar didn't raise no bitch grok-4.7 14/17 didn't even try 'til now, no surprise muse six-seven :( but it's a very good model I swear may be the last one: the attractors are very visible now@zacharynadoRT @tensecorrection: priv8 hacker jeopardy benchmark, saturated edition: gemini-argon-hi 14.5/17 Sundar didn't raise no bitch grok-4…@tensecorrectionpriv8 hacker jeopardy benchmark, saturated edition: gemini-argon-hi 14.5/17 Sundar didn't raise no bitch grok-4.7 14/17 didn't even try 'til now, no surprise muse six-seven :( but it's a very good model I swear may be the last one: the attractors are very visible now@zacharynadoRT @tensecorrection: priv8 hacker jeopardy benchmark, saturated edition: gemini-argon-hi 14.5/17 Sundar didn't raise no bitch grok-4…@tensecorrectionpriv8 hacker jeopardy benchmark, saturated edition: gemini-argon-hi 14.5/17 Sundar didn't raise no bitch grok-4.7 14/17 didn't even try 'til now, no surprise muse six-seven :( but it's a very good model I swear may be the last one: the attractors are very visible now@zacharynadoRT @tensecorrection: priv8 hacker jeopardy benchmark, saturated edition: gemini-argon-hi 14.5/17 Sundar didn't raise no bitch grok-4…—
Not ranked yet
—
Not ranked yet
