Tan-Thinh Duong
tanthinhdt
AI & ML interests
Computer Vision
Recent Activity
updated
a dataset
8 days ago
tanthinhdt/signalp6
published
a dataset
8 days ago
tanthinhdt/signalp6
reacted
to
marksverdhei's
post
with š
8 days ago
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
Organizations
models
4
tanthinhdt/Cytotoxicity-Nanoparticles_Llamma-3.1_20250429-021549
Updated
tanthinhdt/Cytotoxicity-Nanoparticles_BioGPT_20250428-012904
Text Classification
ā¢
0.3B
ā¢
Updated
ā¢
1
tanthinhdt/blip-base_with-pretrained_flickr30k
Image-Text-to-Text
ā¢
0.2B
ā¢
Updated
ā¢
1
tanthinhdt/ViAVSP_LLM_v2_0
Updated