Retrieval Capacity of Self-Attention Under Competition
Researchers measure how few high-attention tokens language models need to match full-context prediction loss.
2 stories, newest first · one card per story
Researchers measure how few high-attention tokens language models need to match full-context prediction loss.
Researchers show repulsive self-attention in a minimal recurrent transformer produces chaos, condensation, and emergent locality.
Today 1.6K readers · 2.5K views · 779 reads · all time 189K views