dolphin-2.9.2-Phi-3-Medium

cognitivecomputations

Capability

percentile index

Downloads

—

Likes

—

Avg score

28.6

6 benchmarks

Weights

Open

Released

—

About

dolphin-2.9.2-Phi-3-Medium is a open-weights model in the dolphin family from cognitivecomputations. Benchmarked on 6 evals, averaging 28.6.

mit

Benchmark scores · 6

avg 28.6 — higher is better

BIG-Bench HardOpen LLM LB · 2026-05

49.7

IFEvalOpen LLM LB · 2026-05

42.5

MMLU-ProOpen LLM LB · 2026-05

39.5

MATH Lvl 5Open LLM LB · 2026-05

18.3

MuSROpen LLM LB · 2026-05

11.4

GPQA DiamondOpen LLM LB · 2026-05

10.3

Details

Familydolphin

Parameters—

Licensemit

WeightsOpen

Released—

History

Capability over time

Append-only ledger — every observed change to this model's numbers.