Sinks are equivalent to hard attention switches that zero out outputs and are cheaper than diagonal patterns when self-communication is allowed, closing the gap between oversmoothing prevention needs and what sinks provide.
Exclusive self attention
4 Pith papers cite this work. Polarity classification is still indexing.
4
Pith papers citing it
citation-role summary
other 1
citation-polarity summary
years
2026 4roles
other 1polarities
unclear 1representative citing papers
A VLA policy with auxiliary success/progress heads and AWR+RECAP-style RL finished 1st in the LeHome 2026 simulation round and 2nd on the real robot.
Belief2-Attention augments prior Belief-Attention with both projected and perpendicular signals plus an extra ZZ^T correlation matrix, claimed more expressive and tested on image classification and segmentation.
citing papers explorer
-
Sink vs. diagonal patterns as mechanisms for attention switch and oversmoothing prevention
Sinks are equivalent to hard attention switches that zero out outputs and are cheaper than diagonal patterns when self-communication is allowed, closing the gap between oversmoothing prevention needs and what sinks provide.
-
Learning to Fold: prizewinning solution at LeHome Challenge 2026 (1st place online, 2nd offline)
A VLA policy with auxiliary success/progress heads and AWR+RECAP-style RL finished 1st in the LeHome 2026 simulation round and 2nd on the real robot.
-
Improved Belief-Attention in Vision Task
Belief2-Attention augments prior Belief-Attention with both projected and perpendicular signals plus an extra ZZ^T correlation matrix, claimed more expressive and tested on image classification and segmentation.
- Do Value Vectors in Deep Layers Need Context from the Residual Stream?