REGER adds inter-item relations to generative recommendation with semantic IDs
The method pretrains by generating semantic neighbors' IDs and distills a teacher's full next-token distributions along the target path.
Stories per day · 30 days
Relatedagent-behavior1arxiv1bias1cuda1distillation1evaluation1generative-recommendation1inference1language-models1llm-agents1preference-distillation1semantic-ids1source-preference1trustworthiness1
6 stories, newest first · one card per story
Stories tagged #recommendation
DuoLM reuses compact item embeddings for faster LLM recommendations
The paper reports about 26,400 candidate scores per second on one H100, roughly 60 times a vLLM-served teacher.
Source Preference in the Wild: How LLM Agents Favor Items by Source, and How to Reduce It
Study of 12 LLM agents finds systematic source preference bias that overrides requirement satisfaction; hiding or relabeling source info changes selection rates.
Decision-Oriented Recommendation Reranking: An Empirical Study of Jev
TypeSafe AI's Jev holds recommendation quality while latency grows more slowly than pointwise Qwen rerankers.
Decision-Oriented Recommendation Reranking: An Empirical Study of Jev
Jev matches strong recommendation rerankers with slower latency growth than pointwise Qwen models.
AutoResearch at Production Scale: Failure Modes and a Multi-Agent Framework
Production AutoResearch on book embeddings hit five failure modes; a scaffolding fix lifted Recall@6 1.82x.