FedDetox uses on-device knowledge-distilled classifiers to sanitize toxic data in federated SLM training, preserving safety alignment comparable to centralized baselines.
PluralLLM: Pluralistic alignment in LLMs via federated learning
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
cs.CR 1years
2026 1verdicts
UNVERDICTED 1representative citing papers
citing papers explorer
-
FedDetox: Robust Federated SLM Alignment via On-Device Data Sanitization
FedDetox uses on-device knowledge-distilled classifiers to sanitize toxic data in federated SLM training, preserving safety alignment comparable to centralized baselines.