Phi-2

Phi-2.

Microsoft Released 12 Dec 2023 Family Phi-2 The lab's model card

Benchmarks with a score

Ranked first, out of five or more models

In the top three

Microsoft Dec 2023

Open weights Small

107.9 #230 of 271

Epoch Capabilities Index

Where it ranks

Its best score on each benchmark, and where that ranks among every model with a score there. We rank each model once, at its best setting.

Benchmark	Score	Rank	Source
OpenBookQA	73.6%	#9 of 41	Epoch AI
ARC-Challenge	75.9%	#20 of 75	Epoch AI
Vectara Hallucination Leaderboard (HHEM)	6.70%	#41 of 148	Official leaderboard
BoolQ	83.3%	#10 of 28	Model card
BIG-Bench Hard	59.2%	#58 of 71	Model card
ANLI	42.5%	#9 of 11	Epoch AI
HellaSwag	53.6%	#67 of 74	Epoch AI
MMLU	56.7%	#117 of 127	Model card
WinoGrande	54.7%	#74 of 76	Epoch AI
EQ-Bench v2	27.6	#83 of 85	Official leaderboard
TriviaQA	45.2%	#38 of 38	Epoch AI
AlpacaEval	30.7%	#19 of 19	Official leaderboard

Benchmark	Score		Rank	Source
GSM8K	61.1%		#74 of 87	Model card

Benchmark	Score	Rank	Source
HumanEval+	45.1%	#33 of 46	Official leaderboard
MBPP+	54.2%	#26 of 35	Official leaderboard
MBPP	59.1%	#53 of 66	Model card
CRUXEval	33.5%	#10 of 11	Official leaderboard

Benchmark	Score		Rank	Source
HELM Lite	16.9		#69 of 71	Official leaderboard

Microsoft's models in color, against every other model on the Epoch Capabilities Index, by release date.

The benchmarks the lab chose to headline, with the scores it gave.