# Keeping the Index Open: The Recommendation-Side Cost of Shared Search and Recommendation

arXiv: https://arxiv.org/abs/2608.24079 · HTML: https://mlsanity.com/paper/2608.24079/

## Key facts

- No citing papers in the map yet (data through 14 Oct 2026).
- Builds on 33 of its 58 references that are in the map.
- First posted on arXiv 25 Aug 2026.

## Builds on (most cited first, 30 of 33)

- [Attention Is All You Need](https://mlsanity.com/paper/1706.03762/) (2017)
- [BERT: Pre-training of Deep Bidirectional Transformers for Language Understanding](https://mlsanity.com/paper/1810.04805/) (2018)
- [Representation Learning with Contrastive Predictive Coding](https://mlsanity.com/paper/1807.03748/) (2018)
- [Billion-scale similarity search with GPUs](https://mlsanity.com/paper/1702.08734/) (2017)
- [Understanding Contrastive Representation Learning through Alignment and Uniformity on the Hypersphere](https://mlsanity.com/paper/2005.10242/) (2020)
- [Self-Attentive Sequential Recommendation](https://mlsanity.com/paper/1808.09781/) (2018)
- [BERT4Rec: Sequential Recommendation with Bidirectional Encoder Representations from Transformer](https://mlsanity.com/paper/1904.06690/) (2019)
- [Efficient and robust approximate nearest neighbor search using Hierarchical Navigable Small World graphs](https://mlsanity.com/paper/1603.09320/) (2016)
- [Session-based Recommendations with Recurrent Neural Networks](https://mlsanity.com/paper/1511.06939/) (2015)
- [Personalized Top-N Sequential Recommendation via Convolutional Sequence Embedding](https://mlsanity.com/paper/1809.07426/) (2018)
- [AutoInt: Automatic Feature Interaction Learning via Self-Attentive Neural Networks](https://mlsanity.com/paper/1810.11921/) (2018)
- [Neural Attentive Session-based Recommendation](https://mlsanity.com/paper/1711.04725/) (2017)
- [Sparse, Dense, and Attentional Representations for Text Retrieval](https://mlsanity.com/paper/2005.00181/) (2020)
- [Are We Really Making Much Progress? A Worrying Analysis of Recent Neural Recommendation Approaches](https://mlsanity.com/paper/1907.06902/) (2019)
- [Embedding-based Retrieval in Facebook Search](https://mlsanity.com/paper/2006.11632/) (2020)
- [RecBole: Towards a Unified, Comprehensive and Efficient Framework for Recommendation Algorithms](https://mlsanity.com/paper/2011.01731/) (2020)
- [Filter-enhanced MLP is All You Need for Sequential Recommendation](https://mlsanity.com/paper/2202.13556/) (2022)
- [Contrastive Learning for Cold-Start Recommendation](https://mlsanity.com/paper/2107.05315/) (2021)
- [Efficient softmax approximation for GPUs](https://mlsanity.com/paper/1609.04309/) (2016)
- [Pixie: A System for Recommending 3+ Billion Items to 200+ Million Users in Real-Time](https://mlsanity.com/paper/1711.07601/) (2017)
- [Cut Your Losses in Large-Vocabulary Language Models](https://mlsanity.com/paper/2411.09009/) (2024)
- [On the Effectiveness of Sampled Softmax Loss for Item Recommendation](https://mlsanity.com/paper/2201.02327/) (2022)
- [Decoupled Side Information Fusion for Sequential Recommendation](https://mlsanity.com/paper/2204.11046/) (2022)
- [Turning Dross Into Gold Loss: is BERT4Rec really better than SASRec?](https://mlsanity.com/paper/2309.07602/) (2023)
- [gSASRec: Reducing Overconfidence in Sequential Recommendation Trained with Negative Sampling](https://mlsanity.com/paper/2308.07192/) (2023)
- [Cross-Batch Negative Sampling for Training Two-Tower Recommenders](https://mlsanity.com/paper/2110.15154/) (2021)
- [When Search Meets Recommendation: Learning Disentangled Search Representation for Recommendation](https://mlsanity.com/paper/2305.10822/) (2023)
- [Scalable Cross-Entropy Loss for Sequential Recommendations with Large Item Catalogs](https://mlsanity.com/paper/2409.18721/) (2024)
- [Beyond Two-Tower Matching: Learning Sparse Retrievable Cross-Interactions for Recommendation](https://mlsanity.com/paper/2311.18213/) (2023)
- [UniSAR: Modeling User Transition Behaviors between Search and Recommendation](https://mlsanity.com/paper/2404.09520/) (2024)

## Abstract (arXiv)

A shared search-and-recommendation index must score new items from features alone because search has no exploration slot. In a public log covering both surfaces over one catalog, 38.6% of held-out query-search impressions show an item never previously shown or visited. For user-cold engagements, the feature-based tower serves this demand without measurable loss against 99 sampled negatives (0.9595 Recall@20 versus 0.9510 warm). A lexical baseline reaches similar parity, while a full-catalog check remains statistically undecided. Dual-encoder retrieval therefore keeps the index open to new items, unlike an ID-softmax recommender that requires retraining. We price this openness on recommendation against six sequential baselines, each retrained and tuned through five rounds on corrected targets. A float32 timestamp bug had reordered leave-one-out targets for 19.7% of users. On MovieLens-1M, warm accuracy trails the strongest retrained baseline by 5.2% Recall@20 and 11.4% NDCG@20. On MIND, the gap narrows to 0.8--3.6% relative to the five strongest baselines, though the model ranks sixth of seven. Under strict zero-leakage cold-start evaluation, the content tower achieves 0.172 ± 0.006 Recall@20, 1.4× the strongest retrained dedicated method (0.124 ± 0.007) and 3× a training-free floor, without cold-specific training. Exact full-softmax training raises Recall@20 by 54% on MIND-small and 6.9% on MovieLens-1M over sampled InfoNCE, but recomputes the full catalog each step and exhausts accelerator memory at 240K items. Approximate nearest-neighbor search explains none of the remaining gap, serving cost does not regress against ID-softmax retrieval, and a history-window sweep explains half the post-recipe remainder. Exact-quality training at catalog scale remains the open problem.