An AI reflector that suggests improvements to AI generated essays. Essays are continually improved until a score of at least 95/100 is reached. Calls fireworks.ai to use different available open source LLMs.
- Test out the efficacy of various LLMs
- Add additional feedback signals in an attempt to get to desired scores more quickly