Research map / Vision-language and multimodal / Cross-modal learning
SynGR: Unleashing the Potential of Cross-Modal Synergy for Generative Recommendation
Wei Chen, Xingyu Guo, Shuang Li, Fuwei Zhang, Meng Yuan, Jing Fan, Zhao Zhang, deqing wang, Fuzhen Zhuang
Read the paperCheck your idea against 137,217 papers
AI-extracted. The measured effects below were read by AI from the paper's full text (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.
Measured effects reported in the paper
| What was changed | Where | Measured | Effect |
|---|---|---|---|
| replace saliency-aware masking with random masking (ablation) | SynGR · Amazon Arts and Games | HR@10, NDCG@10 | Improved |
| replace constructed unimodal negative with in-batch random negatives (ablation) | SynGR · Amazon Arts and Games | HR@10, NDCG@10 | Improved |
| add synergistic contrastive learning objective (ablation) | SynGR · Amazon Arts and Games | HR@10, NDCG@10 | Improved |
| increase masking ratio r to above 0.5 (ablation) | SynGR · Games, Instruments | HR@10, NDCG@10 | Hurt |
| increase contrastive loss weight lambda above 0.005 (ablation) | SynGR · Games | HR@10 | Hurt |
| vary temperature tau in 0.01-0.09 (ablation) | SynGR · Arts | HR@10 | No clear effect |
| use SynGR instead of MACRec | MACRec · Amazon Instruments | HR@10: 0.1363 -> 0.1768 | Improved |
| use SynGR instead of MACRec | MACRec · Amazon Games | HR@10: 0.1078 -> 0.1092 | Improved |
| use SynGR instead of MACRec for training efficiency | MACRec · Arts / Instruments | training time: -23% Arts, -22% Instruments | Improved |
Closest papers by meaning
- Towards Unified Multi-Modal Personalization: Large Vision-Language Models for Generative Recommendation and Beyond (ICLR 2024 · similarity 0.89)
- Actions Speak Louder than Words: Trillion-Parameter Sequential Transducers for Generative Recommendations (ICML 2024 · similarity 0.87)
- Multimodal Quantitative Language for Generative Recommendation (ICLR 2025 · similarity 0.87)
- Semantic Convergence: Harmonizing Recommender Systems via Two-Stage Alignment and Behavioral Semantic Tokenization (AAAI 2025 · similarity 0.86)
- Fine Tuning Out-of-Vocabulary Item Recommendation with User Sequence Imagination (NeurIPS 2024 · similarity 0.86)
- Rethinking Semantic ID Construction for Generative Recommendation: SimHash with Parallel Decoding and Semantic Alignment (NeurIPS 2026 · similarity 0.85)
Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.
