Verdict
Falcon-180B leads 6–2 across 8 shared benchmarks.
Falcon-180B 6 · Llama 2-70B 2 · higher isn't always better — check the caveats below.
Reported scores — protocols may differ. How we compare
Commonsense Reasoning: Falcon-180B leads 3–0 across 3 shared commonsense reasoning benchmarks (largest gap: Winogrande, 74.2 vs 60.4 — single evaluator).
Reasoning: Evenly split 1–1 across 2 shared reasoning benchmarks (largest gap: ARC AI2, 71.07 vs 57.07 — single evaluator).
Falcon-180B leads 6–2 across 8 shared benchmarks. Falcon-180B leads 6 benchmarks and Llama 2-70B leads 2. Higher isn't always better — see the integrity caveats on each benchmark.
Falcon-180B and Llama 2-70B have 8 benchmarks in common in our data — those are the only rows where a direct, apples-to-apples comparison is drawn.
Falcon-180B leads commonsense reasoning 3–0.