Proposes dynamic token re-weighting during target-domain fine-tuning to mitigate exacerbated attention sink in source-free CDFSL, achieving SOTA on four benchmarks.
Decoupling template bias in clip: Harnessing empty prompts for enhanced few-shot learning
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
cs.CV 1years
2026 1verdicts
UNVERDICTED 1representative citing papers
citing papers explorer
-
Addressing Exacerbated Attention Sink for Source-Free Cross-Domain Few-Shot Learning
Proposes dynamic token re-weighting during target-domain fine-tuning to mitigate exacerbated attention sink in source-free CDFSL, achieving SOTA on four benchmarks.