mlsanityData to 14 Oct 2026

RECE: Reduced Cross-Entropy Loss for Large-Catalogue Sequential Recommenders

arXiv 2408.02354 ↗First posted 5 Aug 2024cs.IR cs.LGcited by 9

Not on the map. Why?

Abstract (arXiv)

Scalability is a major challenge in modern recommender systems. In sequential recommendations, full Cross-Entropy (CE) loss achieves state-of-the-art recommendation quality but consumes excessive GPU memory with large item catalogs, limiting its practicality. Using a GPU-efficient locality-sensitive hashing-like algorithm for approximating large tensor of logits, this paper introduces a novel RECE (REduced Cross-Entropy) loss. RECE significantly reduces memory consumption while allowing one to enjoy the state-of-the-art performance of full CE loss. Experimental results on various datasets show that RECE cuts training peak memory usage by up to 12 times compared to existing methods while retaining or exceeding performance metrics of CE loss. The approach also opens up new possibilities for large-scale applications in other domains.

Key facts. Cited by 9 papers in the map (data through 14 Oct 2026). Builds on 26 of its 42 references that are in the map. First posted on arXiv 5 Aug 2024.

Builds on

26 of 42 references.

References · 26

  1. 2017Attention Is All You Needcited by 60,361
  2. 2020Reformer: The Efficient Transformercited by 1,668
  3. 2018Self-Attentive Sequential Recommendationcited by 1,184
  4. 2019BERT4Rec: Sequential Recommendation with Bidirectional Encoder Representations from Transformercited by 964
  5. 2015Image-based Recommendations on Styles and Substitutescited by 706
  6. 2015Session-based Recommendations with Recurrent Neural Networkscited by 695
  7. 2018Personalized Top-N Sequential Recommendation via Convolutional Sequence Embeddingcited by 476
  8. 2017Recurrent Neural Networks with Top-k Gains for Session-based Recommendationscited by 165
  9. 2021Contrastive Learning for Representation Degeneration Problem in Sequential Recommendationcited by 121
  10. 2015Practical and Optimal LSH for Angular Distancecited by 106
  11. 2023Frequency Enhanced Hybrid Attention Network for Sequential Recommendationcited by 52
  12. 2020Exploring Data Splitting Strategies for the Evaluation of Recommendation Modelscited by 48
14 more
  1. 2020A Critical Study on Data Leakage in Recommender System Offline Evaluationcited by 39
  2. 2020Simplify and Robustify Negative Sampling for Implicit Collaborative Filteringcited by 34
  3. 2023Turning Dross Into Gold Loss: is BERT4Rec really better than SASRec?cited by 30
  4. 2015Quantization based Fast Inner Product Searchcited by 29
  5. 2023gSASRec: Reducing Overconfidence in Sequential Recommendation Trained with Negative Samplingcited by 29
  6. 2021A Case Study on Sampling Strategies for Evaluating Neural Sequential Item Recommendation Modelscited by 25
  7. 2021Cross-Batch Negative Sampling for Training Two-Tower Recommenderscited by 24
  8. 2016Vista: A Visually, Socially, and Temporally-aware Model for Artistic Recommendationcited by 20
  9. 2017A New Unbiased and Efficient Class of LSH-Based Samplers and Estimators for Partition Function Computation in Log-Linear Modelscited by 16
  10. 2022Contrastive Learning with Bidirectional Transformers for Sequential Recommendationcited by 10
  11. 2023Scaling Session-Based Transformer Recommendations using Optimized Negative Sampling and Loss Functionscited by 9
  12. 2014Deep Networks With Large Output Spacescited by 7
  13. 2022Tensor-based Sequential Learning via Hankel Matrix Representation for Next Item Recommendationscited by 5
  14. 2022Generating Negative Samples for Sequential Recommendationcited by 3

Built on by

Cited by year (2026 to 14 Oct)

Citing papers per year: 2024: 2; 2025: 3; 2026: 4.

9 papers.

  1. 2024Scalable Cross-Entropy Loss for Sequential Recommendations with Large Item Catalogscited by 15
  2. 2025Time to Split: Exploring Data Splitting Strategies for Offline Evaluation of Sequential Recommenderscited by 12
  3. 2024Autoregressive Generation Strategies for Top-K Sequential Recommendationscited by 5
  4. 2025Recommendation Is a Dish Better Served Warmcited by 2
  5. 2025Maximum Impact with Fewer Features: Efficient Feature Selection for Cold-Start Recommenders through Collaborative Importance Weightingcited by 1
  6. 2026Position-Aware Sequential Attention for Accurate Next Item Recommendationscited by 0
  7. 2026Cross-Representation Knowledge Transfer for Improved Sequential Recommendationscited by 0
  8. 2026Batch Size or Negatives? A Selection Rule for Memory-Constrained Recommender Trainingcited by 0
  9. 2026Keeping the Index Open: The Recommendation-Side Cost of Shared Search and Recommendationcited by 0