@yoheinakajima: longmemeval experiment arch: 1) deterministic ingestion/extraction (85.6% accuracy, 86.2% retrieval) 2) semantic ingest…
Summary
The tweet shares results from LongMemEval experiments comparing deterministic and semantic ingestion/extraction methods, achieving up to 87.6% accuracy and 90.0% retrieval with a hybrid approach.
View Cached Full Text
Cached at: 06/01/26, 05:47 PM
longmemeval experiment arch:
- deterministic ingestion/extraction (85.6% accuracy, 86.2% retrieval)
- semantic ingestion/extraction (84.8% accuracy, 94.9% retention)
- semantic ingestion/deterministic extraction (87.6% accuracy, 90.0% retrieval)
this was staggered, not
ActiveGraph (@ActiveGraphAI): Our third LongMemEval experiments by adding semantic ingestion to deterministic retrieval.
The hybrid approach improved end-to-end QA accuracy to 87.6% and evidence retrieval to 90.0%, but the QA gain did not reach statistical significance in this run.
Similar Articles
@yoheinakajima: i know it's backwards order, but experiment #2:
In our second longmemeval experiment, we introduce semantic ingestion into recall leveraging the ActiveGraph runtime, improving retrieval from 60.6% to 83.4%/84.8% for flat/agentic retrieval with LLM ingestion.
@yoheinakajima: ran my first benchmark this weekend (longmemeval) mostly to test activegraph, learned a lot! - this is a stepping stone…
Yohei Nakajima ran the LongMemEval benchmark on ActiveGraph, achieving 85.6% QA accuracy and 86.2% turn answer-in-context, demonstrating the effectiveness of event-based agent systems for long-term memory.
MemoryOS – AI agent memory with temporal knowledge graph and 9ms ingest and 78ms retrieval
MemoryOS is an open-source, self-hosted AI agent memory tool using a temporal knowledge graph, achieving 86.2% accuracy on LongMemEval-s with fast 78ms retrieval speeds.
Benchmarking agent memory retrieval on LongMemEval‑S — 98% Recall@5, 100% recall by R@23, local embeddings only (all-MiniLM-L6-v2), no LLM, no API key
The author shares benchmark results for memweave, a Python library for agent memory, achieving 98% Recall@5 on LongMemEval-S using only local embeddings without LLM calls. The post details the methodology and compares performance against mempalace, highlighting stable retrieval across different question types.
Memanto: Typed Semantic Memory with Information-Theoretic Retrieval for Long-Horizon Agents
Memanto introduces a typed semantic memory system using a schema, conflict resolution, and Moorcheh's information-theoretic retrieval engine, achieving state-of-the-art results on LongMemEval and LoCoMo benchmarks with zero ingestion cost and sub-90ms latency.