Search papers, labs, and topics across Lattice.
This paper introduces AMBER, a novel event-centric paradigm for LLM-based recommendation systems that encodes rich user, item, context, and outcome signals into compact Event Tokens, enhancing the information captured per interaction. By employing a new scaling dimension called snapshot resolution, AMBER effectively decouples the encoding of temporal snapshots from real-time serving compute, allowing for more efficient processing. The results show that AMBER not only advances the compute-quality Pareto frontier on industrial-scale benchmarks but also outperforms traditional tokenization methods, demonstrating its versatility across different model architectures.
Event Tokens can significantly enhance LLM recommendation systems by capturing rich contextual information, leading to superior performance on industrial benchmarks.
LLM-based recommendation has scaled along model capacity and sequence length, yet each position encodes only text, semantic IDs, or a few categorical features, discarding rich user, item, context, and outcome signals available at each event. Under autoregressive modeling, this yields weak queries at each position and, since each position becomes context for the next, the degradation compounds across the sequence. We propose an event-centric paradigm that represents each interaction by its full temporal snapshot, and identify a new scaling dimension we term snapshot resolution: the amount of information encoded per event. To efficiently scale snapshot resolution, we introduce AMBER (Autoregressive Modeling via Bottlenecked Event Representation), which compresses each temporal snapshot into a compact Event Token, a new LLM input modality. The representation is learned end-to-end, while Event Tokens are pre-computed and cached for serving, decoupling snapshot resolution from real-time serving compute. On industrial-scale ranking and retrieval benchmarks, AMBER advances the compute-quality Pareto frontier relative to alternative recommendation paradigms. At sufficient capacity, a single unified tokenizer even outperforms dedicated per-entity tokenizers, demonstrating positive transfer across structurally different entity types. AMBER's Event Tokens also transfer across model architectures: when integrated into a heavily optimized non-LLM ranker as serving-time historical features, they yield statistically significant improvements. Further scaling Event Tokenizer capacity provides additional improvements.
