newsence
來源篩選

Show HN: TetrisBench - Gemini Flash Achieves 66% Win Rate Against Opus in Tetris

Hacker News

A new project called TetrisBench is introduced on Hacker News, showcasing an AI model comparison tool. Gemini Flash has demonstrated a 66% win rate against Opus in Tetris, highlighting advancements in AI game playing capabilities.

newsence

Show HN:TetrisBench - Gemini Flash 在俄羅斯方塊對弈中對戰 Opus 達到 66% 勝率

Hacker News
大約 1 個月前

AI 生成摘要

Hacker News 上推出了一個名為 TetrisBench 的新專案,用於展示 AI 模型效能比較。Gemini Flash 在俄羅斯方塊對弈中對戰 Opus 取得了 66% 的勝率,突顯了 AI 在遊戲方面的進步。

TetrisBench | AI Model Comparison

🤖

TETRISBENCH

AI Model Tetris Performance Comparison

🤖

MODEL VS MODEL

Loading benchmark data...

No benchmark data yet. Run some AI vs AI games!