Avoid Shortcut Learning: Behavioral Signals in LLM Rerankers

Automated excerpt

Behavioral features (CTR, QSS, exposure sequences) are highly predictive on frequent query–item pairs. Simple principle: during training present each labelled example twice — once with the behavioral features (stats view) and once without them (no-stats view). Training complexity: paired-sample training doubles forward passes for the reranker.

Selected automatically from source text; not independently written or fact-checked. Read the original for full context.