A3 is a learnable activation scaling module that trains on amplified adversarial signals via contrastive losses to improve robustness when the same parameters are used in attenuation mode.
Title resolution pending
2 Pith papers cite this work. Polarity classification is still indexing.
2
Pith papers citing it
years
2026 2verdicts
UNVERDICTED 2representative citing papers
Factorial experiment identifies retriever type, dataset, and retrieval depth as strongest influences on RAG poisoning exposure across 432 configurations.
citing papers explorer
-
Improving Adversarial Robustness via Activation Amplification and Attenuation
A3 is a learnable activation scaling module that trains on amplified adversarial signals via contrastive losses to improve robustness when the same parameters are used in attenuation mode.