You are viewing a single comment's thread from:

RE: LeoThread 2025-01-15 13:38

in LeoFinance β€’ 17 days ago

Coding Section

  • πŸ“ˆ Deepseek V3 performed better in the coding section, with a higher success rate in questions such as generating a snake game and creating a Pac-Man game.
  • πŸ“Š Claude 3.5 Sonnet struggled with some coding questions, including generating a snake game and creating a Pac-Man game, but performed well in others.
  • πŸ“ Both models were able to generate code for a level two question, but Deepseek V3's code was more functional and efficient.