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

arXiv: https://arxiv.org/abs/2602.09401 · HTML: https://mlsanity.com/paper/2602.09401/

## Key facts

- Cited by 3 papers in the map (data through 14 Oct 2026).
- First posted on arXiv 10 Feb 2026.

## Built on by (most cited first, 3 of 3)

- [FLUID: From Ephemeral IDs to Multimodal Semantic Codes for Industrial-Scale Livestreaming Recommendation](https://mlsanity.com/paper/2605.21832/) (2026)
- [SinkRec: Mitigating Semantic State Sink in Long Sequence Recommendation with Memory-Conditioned Gated Delta Networks](https://mlsanity.com/paper/2606.09888/) (2026)
- [Scaling Articulated Rationales for MLLM-based Recommendation](https://mlsanity.com/paper/2609.17639/) (2026)

## Abstract (arXiv)

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 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.