6 ms·
I've found GPT-5 Thinking to perform worse than o3 did in tasks of a similar nature. It makes more bad assumptions that de-rail the train of thought.
by Hugsun 1y ago
I've found GPT-5 Thinking to perform worse than o3 did in tasks of a similar nature. It makes more bad assumptions that de-rail the train of thought.
- 3abiton 1y agoI think the key is prompting, and bound boxing assumptions.