Memory Native Inference

Piramid is an inference systems project investigating how language models can access and use external information directly during computation.

The long term goal is to move beyond prompt-based retrieval by investigating how external information can be represented, retained, and integrated into transformer computation.

from piramid import Piramid

db = Piramid("http://localhost:6333")
db.embed("notes", ["Compaction rewrites the store without deleted documents."])

answer = db.generate(
    messages=[{"role": "user", "content": "What happens to deleted documents?"}],
    collection="notes",
    k=2,
)
print(answer["text"])

In active research. Results and write-ups will be published here.