UTTSI selectively scales test-time compute for CTR prediction by triggering stochastic feature-path exploration only on high-uncertainty instances, yielding gains on four datasets and a 5.3% online CTR lift.
Title resolution pending
3 Pith papers cite this work. Polarity classification is still indexing.
3
Pith papers citing it
years
2026 3verdicts
UNVERDICTED 3representative citing papers
DS-MLP achieves state-of-the-art CTR prediction on three benchmarks using a final vanilla MLP structure trained via knowledge distillation and two alignment strategies.
HeteGenCTR adds learnable per-field difficulty parameters to discrete diffusion pre-training for CTR, driving a self-balancing loss and guided attention that reallocates effort to harder fields and yields gains on benchmarks plus online A/B test.
citing papers explorer
-
Selective Test-Time Compute Scaling for Click-Through Rate Prediction via Uncertainty-Triggered Feature Path Exploration
UTTSI selectively scales test-time compute for CTR prediction by triggering stochastic feature-path exploration only on high-uncertainty instances, yielding gains on four datasets and a 5.3% online CTR lift.
-
Dual-Stream MLP is All You Need for CTR Prediction
DS-MLP achieves state-of-the-art CTR prediction on three benchmarks using a final vanilla MLP structure trained via knowledge distillation and two alignment strategies.
-
Self-Balancing Gradient Allocation for Heterogeneity-Aware Feature Generation in Click-Through Rate Prediction
HeteGenCTR adds learnable per-field difficulty parameters to discrete diffusion pre-training for CTR, driving a self-balancing loss and guided attention that reallocates effort to harder fields and yields gains on benchmarks plus online A/B test.