Unverified50% confidenceFactExact time
Google's TIGER (Transformer Index for Generative Recommenders) proposes representing items with semantic IDs, transforming the recommendation problem into a sequence-to-sequence generation task, and autoregressively generating the semantic IDs of recommended items token by token.
1
Sources
50%
Confidence
Long-term
Relevance
8/17/2026
First Seen
Sources
Netflix GenRec：LLM原生推荐系统如何重塑个性化推荐
Related Claims
UnverifiedCross-Encoders concatenate the query and document into a single input sequence and perform deep interaction modeling through Transformer's full attention mechanism; representative models include Cohere Rerank, BGE-Reranker, and ms-marco-MiniLM.66% similarUnverifiedThe Transformer's encoder is responsible for mapping the input sequence to a contextual representation and is suitable for classification and understanding tasks, while the decoder autoregressively generates the output sequence and is suitable for text generation tasks64% similarVerifiedTransformer-based large language models use attention mechanisms to uniformly encode system prompts, user inputs, and externally retrieved content as homogenized token sequences, making it impossible for the model to distinguish trusted instructions from untrusted data at the semantic level.62% similarUnverifiedThe Transformer architecture generates text in an autoregressive manner, outputting one token at a time and appending it to the input sequence to predict the next token, resulting in a serial inference process61% similarUnverified成熟产品往往组合使用多种方案：先用热词偏置在解码时输出正确术语，再用后处理词典兜底，最后在数据积累后考虑微调60% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/762663API
curl https://kongchang.com/api/v1/knowledge/claims/762663MCP
get_claim(id=762663)