feat: add EntityMemory backend with structured key-value fact extraction#17
Merged
Neal006 merged 1 commit intoMay 24, 2026
Merged
Conversation
EntityMemory extracts (key, value) pairs from messages using regex patterns matching the benchmark's injection/update templates. Facts are stored in a dict and returned as a single system-context block on retrieval. Key properties: - Fact updates overwrite the stored value immediately (no stale value survives) - Token cost scales with unique-fact count, not conversation length - Zero LLM calls, fully deterministic and reproducible Wired into VALID_BACKENDS, _make_memory(), and the CLI --backends flag. Closes Neal006#12
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
2 participants
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
Summary
Adds the
EntityMemorybackend requested in issue #12.Instead of storing raw conversation history,
EntityMemorymaintains a key-value dictionary of extracted facts and returns them as a single compact context block on retrieval.Implementation (
memory/entity.py)"My <key> is <value>.""My <key> has changed to <new_value>."dict[key_name → current_value]; updates overwrite in place, so the stale value is gone immediately.systemmessage:"[Known facts about the user] my city is Mumbai; my age is 28; ...".Properties vs other backends
Wiring
VALID_BACKENDSinevaluation/benchmark.py— added"entity"_make_memory()— dispatches toEntityMemory()main.py—--backendshelp text updatedUsage
Closes #12