Multi-Probe Zero Collision Hash (MPZCH): Mitigating Embedding Collisions and Enhancing Model Freshness in Large-Scale Recommenders

arXiv:2602.17050 · cs.LG · Submitted 2026-02-19 · Read on arXiv

cs.LG

Submitted: 2026-02-19

Updated: 2026-10-06

Terminology

• Multi-Probe Zero Collision Hash — 2× in this paper
• Embeddings — 1× in this paper · explained in 2 episodes
• Linear Probing — 1× in this paper · explained in 2 episodes
• Personalization — 1× in this paper · explained in 2 episodes
• Vector Representations — 1× in this paper · explained in 2 episodes

Sources

• Comparing in context: Improving cosine similarity measures with a metric tensor
• On the Embedding Collapse when Scaling up Recommendation Models
• Learning to Embed Categorical Features without Embedding Tables for Recommendation
• LEARN: Knowledge Adaptation from Large Language Model to Recommendation for Practical Industrial Application
• Monolith: Real Time Recommendation System With Collisionless Embedding Table
• Model Size Reduction Using Frequency Based Double Hashing for Recommender Systems
• Actions Speak Louder than Words: Trillion-Parameter Sequential Transducers for Generative Recommendations
• Unified Embedding: Battle-Tested Feature Representations for Web-Scale ML Systems

Related papers

• Polynomial-Augmented Neural Networks (PANNs) with Weak Orthogonality Constraints for Enhanced Function and PDE Approximation
• AIRL-S: Unifying Reinforcement Learning and Search-Based Test-Time Scaling via Adversarial Inverse Reinforcement Learning
• Transformers as Bayesian In-Context Experimenters: Smoothness-Adaptive Efficient ATE Estimation
• Convergence issues in Relational Concept Analysis based on AOC-posets
• Beliefs Beyond Posteriors: The Predictive Cost of ELBO Binding
• Understanding Diffusion Models via Ratio-Based Function Approximation with SignReLU Networks