Engram - Llama3.1-8b-Insturct - FinQA: leaderboard
Engram v1 A3.T6.9, native task score (%), full reported method roster. Student backbone Llama3.1-8b-Insturct; parameter adaptation and trace-graph retrieval methods remain explicit configurations. Do not transfer adapted scores to the frozen base checkpoint. Source confidence intervals and exact row identities pinned. Held-out performance only, not resource efficiency or admission/retention rates.
Metric: Accuracy / task score (%). Source: arxiv.org. Saturation forecast: Rough model projection: around 2026. 5 models tracked.
Top models
| # | Model | Score | Overall rank |
|---|---|---|---|
| 1 | Llama3.1-8b-Insturct (Engram Run: task=FinQA;method=Engram;effort=source-unreported) | 64.6 | |
| 2 | Llama3.1-8b-Insturct (Engram Run: task=FinQA;method=APOLLO-inspired adaptation;effort=source-unreported) | 63.4 | |
| 3 | Llama3.1-8b-Insturct (Engram Run: task=FinQA;method=MAGE;effort=source-unreported) | 60.8 | |
| 4 | Llama3.1-8b-Insturct (Engram Run: task=FinQA;method=FinQANet-inspired adaptation;effort=source-unreported) | 57.1 | |
| 5 | Llama3.1-8b-Insturct (Engram Run: task=FinQA;method=CoT;effort=source-unreported) | 51.4 |
Interactive version: theaggregate.ai/benchmark?slug=engram-llama3-1-8b-insturct-finqa · How It Works · Data refreshed daily, snapshot 2026-11-12.