Randomized layer shuffling and column-wise sign flips make interfering task deltas nearly orthogonal, allowing one compressed superset to retrieve near-fine-tuned accuracy for many tasks.
On the efficacy of knowledge distillation
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
cs.LG 1years
2025 1verdicts
CONDITIONAL 1representative citing papers
citing papers explorer
-
RanDeS: Randomized Delta Superposition for Multi-Model Compression
Randomized layer shuffling and column-wise sign flips make interfering task deltas nearly orthogonal, allowing one compressed superset to retrieve near-fine-tuned accuracy for many tasks.