Different result for similar query

[Crispae]

Result 1. (@Crispae)

Result 2. (@deepika060193 )

Reducing the temperature might help to generate a similar result. Refer to the following discussion. As generating cypher query is similar to coding, we don't want the model to be creative, setting top-P and temperature to 0.1, might work as suggested the in this discussion

[Crispae]

Discussion on, How to regulate the response of OpenAI model

https://community.openai.com/t/cheat-sheet-mastering-temperature-and-top-p-in-chatgpt-api-a-few-tips-and-tricks-on-controlling-the-creativity-deterministic-output-of-prompt-responses/172683

[Crispae]

In the new update of OPENAI model, they released seed parameter, which up to certain extent can provide deterministic results.
https://platform.openai.com/docs/guides/text-generation/reproducible-outputs

[Crispae]

In production version (deployed website), parameters temperature and top_p is changed to 0.2 and 0.1 respectively. In development, it's good idea to release a config file for these parameters.