Updated daily at 05:00 UTC
- Released
- 5 Nov 2019
- Status
- Legacy
- Context
- 1K
- Max output
- —
- Input per 1M tokens
- —
- Output per 1M tokens
- —
- Open weights
- Yes
- License
- modified-mit
- Parameters
- 1.5B
- Knowledge cutoff
- —
- Reasoning
- No
- Modalities
- text → text
Category scores
Out of 100, from independent and third-party results only, by method category-2026-09-29.1.
Not enough independent results yet
Only independent and third-party results rank a model; self-reported and unattributed numbers never do.
Benchmark results
Every result we hold for this model, with who ran it and where it was published.
Knowledge & reasoning
| Benchmark | Metric | Value | Provenance | Setting | Evaluated | Source |
|---|---|---|---|---|---|---|
| AI2 Reasoning Challenge (ARC) | Challenge set accuracy | 25.0% | Self-reported | — | — | Textbooks Are All You Need II: phi-1.5 technical report via Epoch AI |
| HellaSwag | Accuracy | 40.0% | Self-reported | — | — | Textbooks Are All You Need II: phi-1.5 technical report via Epoch AI |
| OpenBookQA | Accuracy | 22.4% | Self-reported | — | — | Textbooks Are All You Need II: phi-1.5 technical report via Epoch AI |
| PIQA | Accuracy | 70.5% | Self-reported | — | — | Textbooks Are All You Need II: phi-1.5 technical report via Epoch AI |
| WinoGrande | Accuracy | 58.3% | Self-reported | — | — | Textbooks Are All You Need II: phi-1.5 technical report via Epoch AI |
| WinoGrande | Accuracy | 58.3% | Source not stated | — | — | Epoch AI via Epoch AI |
Coding & software engineering
| Benchmark | Metric | Value | Provenance | Setting | Evaluated | Source |
|---|---|---|---|---|---|---|
| METR Time Horizons | Average task success (Epoch's column) | 10.1% | Third party | — | — | https://metr.org/blog/2025-03-19-measuring-ai-ability-to-complete-long-tasks/ via Epoch AI |
Other GPT-2 versions
- GPT-2 Large (774M)20 Aug 2019
- GPT-2 Medium (355M)May 2019
- GPT-2 Small (124M)14 Feb 2019