Verdict
Claude 2 leads 4–1 across 5 shared benchmarks.
Claude 2 4 · Llama 2-70B 1 · higher isn't always better — check the caveats below.
Reported scores — protocols may differ. How we compare
Knowledge: Evenly split 1–1 across 2 shared knowledge benchmarks (largest gap: MMLU, 71.33 vs 59.87 — single evaluator).
Math: Claude 2 leads 2–0 across 2 shared math benchmarks (largest gap: MATH level 5, 11.73 vs 3.29 — single evaluator).
Claude 2 leads 4–1 across 5 shared benchmarks. Claude 2 leads 4 benchmarks and Llama 2-70B leads 1. Higher isn't always better — see the integrity caveats on each benchmark.
Claude 2 and Llama 2-70B have 5 benchmarks in common in our data — those are the only rows where a direct, apples-to-apples comparison is drawn.
Claude 2 leads math 2–0.