3 points | by pongogogo 15 hours ago ago
1 comments
Super interesting to see how - for the same set of prompts - GPT and Claude rely on just 50-60 domains while Perplexity and Gemini cite 100-140.
Super interesting to see how - for the same set of prompts - GPT and Claude rely on just 50-60 domains while Perplexity and Gemini cite 100-140.