Giuliano Giacaglia 🌲
@giu
It is surprising how well Gemini 2.5 Pro does in many benchmarks but not talked about a lot. I wonder why there isn't as much attention in the Google model. Here is a view of how most models perform on a PhD-level science reasoning benchmark called GPQA Diamond
4 replies
2 recasts
21 reactions
Garrett
@garrett
long time between updates relative to others and 2.5 has only been around for 2 months
1 reply
0 recast
1 reaction
Gabriel Ayuso
@gabrielayuso.eth
I've switched to using it exclusively for coding since it's significantly better than the rest even Claude
0 reply
0 recast
1 reaction
shoni.eth
@alexpaden
it’s definitely picked up the last few months but the benchmarks just haven’t usually tracked with real world experience imo
0 reply
0 recast
1 reaction
yekta ✎
@yekta
Marketing
0 reply
0 recast
0 reaction