4 ms·
The leader, perhaps, but not by a large margin, and only on these sample benchmarks. "Co-state-of-the-art" is the term used in the article, and I'm going to tak
by bccdee 2y ago
The leader, perhaps, but not by a large margin, and only on these sample benchmarks. "Co-state-of-the-art" is the term used in the article, and I'm going to take that at face value.
- starspangled 2y agoIt significantly outperformed competitors on those benchmarks. Around as much as the deltas between some others, which are considered significant.
- bccdee 2y agoThe deltas between the others are mostly not significant either. They're all about equally good. There's no categorical difference between GPT-4 and Claude 3.5.
- starspangled 2y agoThat's not true.
- bccdee 2y agoOkay what's the categorical difference? Which meaningful category includes one but not the other?