A generative agent scores each memory by combining recency, importance and relevance to the current situation. Equal weights are the simplest choice, and they are what the paper uses after normalizing each term.
Implement retrieval_score(recency, importance, relevance), the equally weighted average.
Add the three values and divide by three.
No results yet
Run your code to check it against the examples, or submit to run every test.