UniPinRec: Unifying Generative Retrieval and Ranking at Pinterest Scale
cs.IR, cs.LG
Submitted: 2026-05-29
Updated: 2026-05-29
Journal ref: Proceedings of the 20th ACM Conference on Recommender Systems (RecSys '26), 2026
License: http://creativecommons.org/licenses/by/4.0/
The gist: Modern recommendation systems predominantly train retrieval and ranking as separate models despite both increasingly relying on large transformers encoding the same user behavior data, duplicating
Terminology
Abstract
Modern recommendation systems predominantly train retrieval and ranking as separate models despite both increasingly relying on large transformers encoding the same user behavior data, duplicating parameters, compute, and serving cost. Prior work unifies the model architecture but not the full pipeline: input formats, training procedures, and serving stacks remain fragmented across stages. We present UniPinRec, which achieves full-stack unification of retrieval and ranking at Pinterest: one input format, one model, one training stage, deployed within existing serving infrastructure. A shared transformer encodes the user action sequence into candidate-independent representations that branch into retrieval (ANN dot-product) and ranking (cross-attention) via task-specific heads. Three ideas make this work: (1) Masked Action Modeling (MAM) eliminates interleaving, enabling weight sharing without doubling context length; (2) Blended training examples pair action sequences with feedview impression slates to satisfy both objectives jointly; (3) Cross-stage KV cache sharing reuses user-history computation from retrieval for ranking, reducing total FLOPs versus serving two independent models. Deployed in the Pinterest core surfaces, UniPinRec delivers approximately +1% online engagement lift while cutting end-to-end serving latency by 11.1% and lifting QPS by 63.6%. To our knowledge, this is the first full-stack unification of retrieval and ranking, covering inputs, model, training and serving, deployed in a production recommendation system.
Sources
- PinRec: Unified Generative Retrieval for Pinterest Recommender Systems
- PinCLIP: Large-scale Foundational Multimodal Representation at Pinterest
- UniSearch: Rethinking Search System with a Unified Generative Architecture
- OnePiece: Bringing Context Engineering and Reasoning to Industrial Cascade Ranking System
- OneRec: Unifying Retrieve and Rank with Generative Recommender and Iterative Preference Alignment
- How Well Does Generative Recommendation Generalize?
- SynerGen: Contextualized Generative Recommender for Unified Search and Recommendation
- Flex Attention: A Programming Model for Generating Optimized Attention Kernels
- Towards Large-scale Generative Ranking
- CADET: Context-Conditioned Ads CTR Prediction With a Decoder-Only Transformer
- Vectorizing the Trie: Efficient Constrained Decoding for LLM-based Generative Retrieval on Accelerators
- OneRanker: Unified Generation and Ranking with One Model in Industrial Advertising Recommendation
- RelayGR: Scaling Long-Sequence Generative Recommendation via Cross-Stage Relay-Race Inference
- IntSR: An Integrated Generative Framework for Search and Recommendation
- Unifying Generative and Dense Retrieval for Sequential Recommendation
- FuXi-\beta: Towards a Lightweight and Fast Large-Scale Generative Recommendation Model
- GPR: Towards a Generative Pre-trained One-Model Paradigm for Large-Scale Advertising Recommendation
- SMES: Towards Scalable Multi-Task Recommendation via Expert Sparsity
Related papers
- The Price of Isolation: Estimating the Ecosystem Cost of Symmetric Two-Sided A/B Testing
- SCAR: Semantic Continuity-Aware Retrieval for Efficient Context Expansion in RAG
- MixLoRA-DSI: Dynamically Expandable Mixture-of-LoRA Experts for Rehearsal-Free Generative Retrieval over Dynamic Corpora
- How Do LLM-Generated Texts Impact Term-Based Retrieval Models?
- MemSyco-Bench: Benchmarking Sycophancy in Agent Memory
- Training with Missed Targets in Generative Recommendation: Separating Supervision from Probability Competition