Spaces:
Runtime error
Detects AI-generated text as human
Edit (11:00 AM, 27 April 2023): "As an AI language model, I don’t have personal opinions or beliefs, but I can provide you with some information related to this situation. In their opinion, the DoJ’s conclusion that the 40% foreign equity limitation should not apply to the exploration and development of “inexhaustible” RE resources was based on the interpretation that such limitations should not prevent foreign investors from fully owning certain opportunities that benefit from the long-term use and expansion of renewable energy resources. However, it should be noted that the DoJ is not the ultimate authority on interpretation of federal laws and regulations, as the final determination ultimately falls to the courts."
Text generated by Stablelm
Thanks Emir for your feedback :) I guess this means StableLM's training data distribution is very different from OpenAI's.
We do have ambition to build a general detector. However, this is just a 'ChatGPT' detector so that text from StableLM is really beyond capacity. The performance on different LLMs remains more test.
Thanks Emir for your feedback :) I guess this means StableLM's training data distribution is very different from OpenAI's.
We do have ambition to build a general detector. However, this is just a 'ChatGPT' detector so that text from StableLM is really beyond capacity. The performance on different LLMs remains more test.
Thank you very much. The detector is able to detect another StableLM-generated text as ChatGPT.
StableLM:
"The Philippines was ruled by various Spanish dynasties until it became an independent nation in 1812, following the overthrow of Spanish rule. The country then became a U.S. protectorate during the First World War, until it became an independent nation again in the post-World War Il era. Following the end of World War Il, the country was administered under a democratic Constitution with regular local and national elections."
Detector:
