3 ms·

https://github.com/BICLab/SpikingBrain-7B/blob/main/assets/table1.png https://github.com/BICLab/SpikingBrain-7B/blob/main/assets/t... Shouldn’t one bold the be

https://github.com/BICLab/SpikingBrain-7B/blob/main/assets/table1.png https://github.com/BICLab/SpikingBrain-7B/blob/main/assets/t... Shouldn’t one bold the better numbers?

rpunkfu 1y ago
Inspired by GPT-5 presentation :)
- doph 1y ago
  Did they ever address that? I have not been able to stop thinking about it, it was so bizarre.
  - rpunkfu 1y ago
    No idea, didn't follow the news on it afterwards.
  - andy_ppp 1y ago
    I honestly thing they just styled it out as "more efficient" and their investors lapped it up. We've really not seen improvement since GPT4 for most things I do with is and it feels like a downgrade on o3-mini-high to me.
daveguy 1y ago
Well, then none of their model's numbers would be bold and that's not what they/AIs usually see in publications!
- cubefox 1y ago
  They do look pretty good compared to the two other linear (non-Transformer) models. Conventional attention is hard to beat in benchmarks but it is quadratic in time and memory complexity.