4 ms·
Hasn't nailed the strawberry test yet
by mudkipdev 1y ago
Hasn't nailed the strawberry test yet
- pxc 1y agoI found this surprising because that's such an old test that it must certainly be in the training data. I just tried to reproduce and I've been unable to get it (20B model, lowest "reasoning" budget) to fail that test (with a few different words).
- quatonion 1y agoI am starting to get the impression the strawberry test is an OpenAI watermark, more than an actual problem. It is a good way to detect if another model was trained on your data for example, or is a distillation/quant/ablation.