MAPE RQS Transplant Rerun - LongMemEval Knowledge Update: leaderboard

MAPE evaluates frozen GPT-4.1 Mini and Qwen3-30B-A3B-Instruct-2507 with different memory programs. LoCoMo uses 1,307 evaluation questions and LongMemEval-S uses 400, with question-weighted accuracy. The evaluation partition also selects the final memory program. The RQS transplant experiment is a separate rerun. This board reports lLM-judged correctness (%). Higher is better.

Metric: LLM-judged correctness (%). Source: arxiv.org. Saturation forecast: Rough model projection: around 2026. 1 model tracked. Updated 14 Nov 2026.

Top models

#ModelSettingScoreOverall rank
1Qwen3-30B-A3B-Instruct-2507Evaluation run: MAPE RQS transplant LightMem with90.48

Interactive version: theaggregate.ai/benchmark?slug=mape-rqs-transplant-rerun-longmemeval-knowledge-update · How It Works · Data refreshed daily, snapshot 2026-11-16.