Engram - Qwen3-8B - WebShop: leaderboard
Engram v1 S5.T2.6.2, native task performance (%), full reported method roster. Student backbone Qwen3-8B; trained adapters and memory frameworks remain native run configurations, not released-model aliases. Full table and confidence intervals retained; no resource, conditional verification or retention rate admitted.
Metric: Accuracy / task score (%). Source: arxiv.org. Saturation forecast: Rough model projection: around 2026. 10 models tracked.
Top models
| # | Model | Score | Overall rank |
|---|---|---|---|
| 1 | Qwen3-8B (Engram Run: method=MAGE;effort=source-unreported) | 84 | |
| 2 | Qwen3-8B (Engram Run: method=Engram;effort=source-unreported) | 83.4 | |
| 3 | Qwen3-8B (Engram Run: method=EDGE;effort=source-unreported) | 82.6 | |
| 4 | Qwen3-8B (Engram Run: method=SkillRL;effort=source-unreported) | 72.7 | |
| 5 | Qwen3-8B (Engram Run: method=GRPO;effort=source-unreported) | 66.1 | |
| 6 | Qwen3-8B (Engram Run: method=ReAct;effort=source-unreported) | 40 | |
| 7 | Qwen3-8B (Engram Run: method=Reflexion;effort=source-unreported) | 28.8 | |
| 8 | Qwen3-8B (Engram Run: method=EvolveR;effort=source-unreported) | 17.6 | |
| 9 | Qwen3-8B (Engram Run: method=MemRL;effort=source-unreported) | 9.2 | |
| 10 | Qwen3-8B (Engram Run: method=Mem0;effort=source-unreported) | 2 |
Interactive version: theaggregate.ai/benchmark?slug=engram-qwen3-8b-webshop · How It Works · Data refreshed daily, snapshot 2026-11-12.