Model comparison
Phi-1.5 vs Phi-2
Neither Phi-1.5 nor Phi-2 has enough public benchmark results to be ranked yet; the rows below show what has been published.
Last verified . 6 shared benchmarks.
MicrosoftSummary
- They share 6 benchmarks with published results for both.
Side by side
| Phi-1.5 | Phi-2 | |
|---|---|---|
| Provider | Microsoft | Microsoft |
| Noometry Index | — | — |
| Released | 2023-09-11 | 2023-12-12 |
| Weights | Open | Open |
| Context window | — | — |
| Max output | — | — |
| Input $ / M tokens | — | — |
| Output $ / M tokens | — | — |
| Results tracked | 7 | 11 |
Sponsored placements are available on pages like this one. Advertise on Noometry
Category by category
Coding Not comparable
Phi-1.5: —, Phi-2: —
| Benchmark | Phi-1.5 | Phi-2 |
|---|---|---|
| HumanEval+ | — | 45.1% |
| MBPP+ | — | 54.2% |
Reasoning Not comparable
Phi-1.5: —, Phi-2: —
| Benchmark | Phi-1.5 | Phi-2 |
|---|---|---|
| Epoch Capabilities Index | 90.82 | 107.62 |
| HellaSwag | 47.6% | 53.6% |
| WinoGrande | 73.4% | 54.7% |
| Adversarial NLI | — | 42.5% |
| BIG-Bench Hard | — | 59.4% |
Knowledge Not comparable
Phi-1.5: —, Phi-2: —
| Benchmark | Phi-1.5 | Phi-2 |
|---|---|---|
| ARC (AI2) Challenge | 44.4% | 75.9% |
| MMLU | 37.6% | 58.4% |
| OpenBookQA | 37.2% | 73.6% |
| BoolQ | 75.8% | — |
| TriviaQA | — | 45.2% |
Frequently asked questions
Is Phi-1.5 better than Phi-2?
Neither Phi-1.5 nor Phi-2 has enough public benchmark results to be ranked yet; the rows below show what has been published.
How many benchmarks do Phi-1.5 and Phi-2 share?
6 benchmarks have published results for both models. Phi-1.5 has 7 scored results on Noometry and Phi-2 has 11.