SoftVQ-VAE: Efficient 1-Dimensional Continuous Tokenizer
TiTok 框架 + Softmax 版 Vector Quantization 以期更高的压缩比
2025
EARN: Efficient Inference Acceleration for LLM-based Generative Recommendation by Register Tokens
通过减小 KV cache size 加速 LLMRec 的推理
Sparse Meets Dense: Unified Generative Recommendations with Cascaded Sparse-Dense Representations
COBRA, 链接离散编码和稠密表示的尝试
OneRec: Unifying Retrieve and Rank with Generative Recommender and Preference Alignment
OneRec, 端到端的推荐模型
Restructuring Vector Quantization with The Rotation Trick
一种利用 Rotation Trick 来替代 STE 的方案
Is Every Item Worth An Embedding?
是否每个 Item 都值得一个可学习的 Embedding 呢
Simba: 符号梯度的可行性
在 Lion 优化器之上的一些探索
Round and Round We Go! What makes Rotary Positional Encodings useful?
理解 RoPE 的高低频
Data Augmentation as Free Lunch: Exploring the Test-Time Augmentation for Sequential Recommendation
TTA, Test-Time Augmentation
Language Representations Can be What Recommenders Need: Findings and Potentials
Next-token embedding 之于协同过滤