1
Companies are optimizing models for specific benchmarks
两大AI巨头为争夺特定基准榜首各自优化模型，GPT-5.6与Opus 5各擅胜场，竞争白热化。
Openai is optimizing for gpqa diamond and anthropic is optimizing for humanity last exam. gpt 5.6 wins on gpqa and opus 5 wins on humanity last exam C…