works on

From the 1 of 19 linked papers with an AI index.

activity
20242026
collaborators
Showing cs.IRShow all

8 papers · 1 filter

cs.IR2026

The Pitfall of Scaling Up: Uncovering and Mitigating Popularity Bias Amplification in Scaling Transformer-based Recommenders

Weiqin Yang, Yue Pan, Chongming Gao +4

We identify a critical pitfall in scaling transformer-based sequential recommenders: while increasing model size improves recommendation accuracy, it simultaneously amplifies popul…

cs.IR2026

Trie-Aware Transformers for Generative Recommendation

Zhenxiang Xu, Jiawei Chen, Sirui Chen +5

Generative recommendation (GR) aligns with advances in generative AI by casting next-item prediction as token-level generation rather than score-based ranking. Most GR methods adop…

cs.IR2026

Talos: Optimizing Top- Accuracy in Recommender Systems

Shengjia Zhang, Weiqin Yang, Jiawei Chen +5

Recommender systems (RS) aim to retrieve a small set of items that best match individual user preferences. Naturally, RS place primary emphasis on the quality of the Top- result…

cs.IR2026

TopKGAT: A Top-K Objective-Driven Architecture for Recommendation

Sirui Chen, Jiawei Chen, Canghong Jin +4

Recommendation systems (RS) aim to retrieve the top-K items most relevant to users, with metrics such as Precision@K and Recall@K commonly used to assess effectiveness. The archite…

cs.IR2025

Breaking the Top- Barrier: Advancing Top- Ranking Metrics Optimization in Recommender Systems

Weiqin Yang, Jiawei Chen, Shengjia Zhang +5

In the realm of recommender systems (RS), Top- ranking metrics such as NDCG@ are the gold standard for evaluating recommendation performance. However, during the training of…

cs.IR2025

Advancing Loss Functions in Recommender Systems: A Comparative Study with a Rényi Divergence-Based Solution

Shengjia Zhang, Jiawei Chen, Changdong Li +5

Loss functions play a pivotal role in optimizing recommendation models. Among various loss functions, Softmax Loss (SL) and Cosine Contrastive Loss (CCL) are particularly effective…