REVIEW 1 cited by
Shapley Head Pruning: Identifying and Removing Interference in Multilingual Transformers
Not yet reviewed by Pith; the record is open.
This paper has not been read by Pith yet. Machine review is queued; the pith claim, tier, and objections will appear here once it completes.
SPECIMEN: schema-true, not a live event
T0 review · schema-true
One-sentence machine reading of the paper's core claim.
pith:XXXXXXXX · record.json · timestamp
read the original abstract
Multilingual transformer-based models demonstrate remarkable zero and few-shot transfer across languages by learning and reusing language-agnostic features. However, as a fixed-size model acquires more languages, its performance across all languages degrades, a phenomenon termed interference. Often attributed to limited model capacity, interference is commonly addressed by adding additional parameters despite evidence that transformer-based models are overparameterized. In this work, we show that it is possible to reduce interference by instead identifying and pruning language-specific parameters. First, we use Shapley Values, a credit allocation metric from coalitional game theory, to identify attention heads that introduce interference. Then, we show that removing identified attention heads from a fixed model improves performance for a target language on both sentence classification and structural prediction, seeing gains as large as 24.7\%. Finally, we provide insights on language-agnostic and language-specific attention heads using attention visualization.
Forward citations
Cited by 1 Pith paper
-
Assortment of Attention Heads: Accelerating Federated PEFT with Head Pruning and Strategic Client Selection
A federated fine-tuning method prunes 90% of attention heads, weights updates by attention importance, and selects clients by loss gap, cutting communication 1.8x and training compute 3.9x with under 2% accuracy drop.
Discussion (0). Continue with ORCID to comment.