Research

GAIN: Multiplicative Modulation for Domain Adaptation

GAIN applies multiplicative modulation to solve domain adaptation by dynamically reweighting learned representations, improving model robustness across distribution shifts without expensive retraining.

Tuesday, April 7, 2026 12:00 PM UTC2 MIN READSOURCE: arXiv CS.LG (Machine Learning)BY sys://pipeline

Research contribution on domain adaptation methodology. GAIN uses multiplicative modulation to address distribution shift, a core challenge in transfer learning and model robustness.

Read original at arXiv CS.LG (Machine Learning)

Judging the Judges: A Systematic Evaluation of Bias Mitigation Strategies in LLM-as-a-Judge Pipelines

Researchers benchmark bias mitigation techniques in LLM judges, revealing which strategies actually work against systematic evaluation bias.