SARM: LLM-Augmented Semantic Anchor for End-to-End Live-Streaming Ranking

TL;DR

SARM introduces an end-to-end ranking model that integrates natural-language semantic anchors with multimodal content for improved live-streaming recommendation accuracy, addressing limitations of previous methods.

Contribution

The paper presents SARM, a novel architecture that jointly optimizes semantic anchors with ranking features, enabling fine-grained, content-aware recommendations in real-time.

Findings

Consistent offline performance improvements.

Successful large-scale A/B testing results.

Deployed at scale serving over 400 million users daily.

Abstract

Large-scale live-streaming recommendation requires precise modeling of non-stationary content semantics under strict real-time serving constraints. In industrial deployment, two common approaches exhibit fundamental limitations: discrete semantic abstractions sacrifice descriptive precision through clustering, while dense multimodal embeddings are extracted independently and remain weakly aligned with ranking optimization, limiting fine-grained content-aware ranking. To address these limitations, we propose \textbf{SARM}, an end-to-end ranking architecture that integrates natural-language semantic anchors directly into ranking optimization, enabling fine-grained author representations conditioned on multimodal content. Each semantic anchor is represented as learnable text tokens jointly optimized with ranking features, allowing the model to adapt content descriptions to ranking objectives. A lightweight dual-token gated design captures domain-specific live-streaming semantics, while an asymmetric deployment strategy preserves low-latency online training and serving. Extensive offline evaluation and large-scale A/B tests show consistent improvements over production baselines. SARM is fully deployed and serves over 400 million users daily.