https://cdn.openai.com/gpt-4-5-system-card.pdf?utm_source=chatgpt.com
SWE-bench得38%，睇來coding都係不及Claude
感覺都係incremental upgrade
ChatGPT-4.5 system card
大棍巴
8 回覆
5 Like
1 Dislike
system card:
https://cdn.openai.com/gpt-4-5-system-card.pdf?utm_source=chatgpt.com
SWE-bench得38%，睇來coding都係不及Claude
感覺都係incremental upgrade
https://cdn.openai.com/gpt-4-5-system-card.pdf?utm_source=chatgpt.com
SWE-bench得38%，睇來coding都係不及Claude
感覺都係incremental upgrade
Processed by a python script generated by LLM in 5 mins
Introducing GPT-4.5
A research preview of our strongest GPT model. Available to Pro users and developers worldwide.
https://openai.com/index/introducing-gpt-4-5/
SimpleQA measures LLM (large language model) factuality on straightforward but challenging knowledge questions.
A research preview of our strongest GPT model. Available to Pro users and developers worldwide.
https://openai.com/index/introducing-gpt-4-5/
SimpleQA measures LLM (large language model) factuality on straightforward but challenging knowledge questions.
真係等下星期試過先知
Grok3出左我已經無用過ChatGPT
grok-3 同 chatGPT o3 Claude Sonnie 3.7邊個好用啲