3 ms·
The logical reasoning in Bard is off in this example (as its basically just putting each destination together with a list of attractions in each city). Additio
by htrp 4y ago
The logical reasoning in Bard is off in this example (as its basically just putting each destination together with a list of attractions in each city).
Additional prompting to tell it to correct the output doesn't seem to have the same effect as it does in the OpenAI models.
- danpalmer 4y agoSorry, can you be more specific? As far as I can tell both are a list of destinations, activities in each, and some minor facts to motivate visits. Arguably there's an issue in the Bard one in that it splits the Tokyo portion in two and has you return there to fly home, but also some people might prefer this to have a round-trip ticket to the same airport. Maybe it's my lack of knowledge about Japan, but I'm just not seeing a difference between the two in terms of logical reasoning.