hybridfree/eevblog-posts
Viewer
ā¢ Updated
ā¢ 207k ā¢ 7
Seven
hybridfree
Ā·
AI & ML interests
None yet
Recent Activity
published
a dataset 17 days ago
hybridfree/eevblog-posts new activity
17 days ago
huggingface/InferenceSupport:hybridfree/evomind_sera14b_unsloth updated
a model 17 days ago
hybridfree/Svene-14b Organizations
None yet
published a dataset 17 days ago
hybridfree/evomind_sera14b_unsloth
#7970 opened 17 days ago
by
hybridfree
updated a model 17 days ago
published a model 17 days ago
reacted to marksverdhei's post with š 20 days ago
Post
4571
Poll: Will 2026 be the year of subquadratic attention?
The transformer architecture is cursed by its computational complexity.
It is why you run out of tokens and have to compact. But some would argue that this is a feature not a bug and that this is also why these models are so good. We've been doing a lot of research on trying to make equally good models that are computationally cheaper, But so far, none of the approaches have stood the test of time. Or so it seems.
Please vote, don't be shy. Remember that the Dunning-Kruger effect is very real, so the person who knows less about transformers than you is going to vote. We want everyone's opinion, no matter confidence.
š if you think at least one frontier model* will have no O(n^2) attention by the end of 2026
š„ If you disagree
* Frontier models - models that match / outperform the flagship claude, gemini or chatgpt at the time on multiple popular benchmarks
The transformer architecture is cursed by its computational complexity.
It is why you run out of tokens and have to compact. But some would argue that this is a feature not a bug and that this is also why these models are so good. We've been doing a lot of research on trying to make equally good models that are computationally cheaper, But so far, none of the approaches have stood the test of time. Or so it seems.
Please vote, don't be shy. Remember that the Dunning-Kruger effect is very real, so the person who knows less about transformers than you is going to vote. We want everyone's opinion, no matter confidence.
š if you think at least one frontier model* will have no O(n^2) attention by the end of 2026
š„ If you disagree
* Frontier models - models that match / outperform the flagship claude, gemini or chatgpt at the time on multiple popular benchmarks
- 4 replies
liked a model about 2 months ago
updated a model about 2 months ago
Reviews and Suggestions
#1 opened about 2 months ago
by
hybridfree
updated a model about 2 months ago
published a model about 2 months ago
liked a model about 2 months ago
updated a dataset about 2 months ago
published a dataset about 2 months ago
published a model about 2 months ago
updated a dataset about 2 months ago
published a dataset about 2 months ago
updated a dataset about 2 months ago
published a dataset about 2 months ago