LATTE: Forecasting Peer Anchored Preference Trajectories for Personalized LLM Generation

2026-05-26

AI Summary

This work addresses the limitations of existing personalization methods for large language models, which compress user history into static representations and fail to disentangle stable preferences, recent shifts, and item-specific content, thereby missing dynamic preference evolution. The authors propose LATTE, a novel framework that constructs relative preference states by anchoring a user’s responses to an item against those of similar users, modeling these states as predictable temporal trajectories. Personalization is achieved by injecting a single soft token—derived from this trajectory—into a frozen large language model. LATTE introduces a peer-anchoring mechanism that separates shared item effects from individual differences and uniquely treats preferences as dynamic trajectories rather than static snapshots. Experiments on Amazon Reviews 2023 and MemoryCD show that LATTE outperforms static latent-variable baselines (ROUGE-L: 0.219) and the strongest compression-based approach (0.245), achieving a ROUGE-L score of 0.259, thus validating the efficacy of trajectory-based preference modeling.

Abstract

Personalized generation with frozen large language models requires a conditioning signal that is both compact and current. Existing personalization methods typically retrieve or summarize user histories in text, or compress them into static latent profiles and soft prompts. These approaches are efficient, but they treat a user's past behavior as an aggregate profile and therefore mix stable identity, recent drift, and item content in the same representation. We propose LAtent Trajectory Tracking and Extrapolation (LATTE), a framework that represents personalization as forecasting a peer anchored relative preference state. For each historical session, LATTE subtracts a time masked baseline formed from comparable users who responded to the same item, producing a state that measures how the target user differs from peers under a shared item context. A lightweight sequence predictor then forecasts the next state in this trajectory, and a State to Token Bridge injects the forecast into a frozen instruction tuned LLM through a single anchored soft token. We provide a latent factor analysis showing when peer anchoring cancels shared item variation and why temporal forecasting trades off stale averages against noisy recent states. Experiments on Amazon Reviews 2023 and MemoryCD show that LATTE consistently outperforms retrieval, summary memory, static latent profiles, difference aware latent profiles, and soft prompt compression baselines. On Amazon Reviews 2023, LATTE improves average ROUGE-L from 0.219 for a static latent profile and 0.245 for the strongest added latent compression baseline to 0.259. Additional pairwise comparisons and diagnostic analyses suggest that the improvement is mainly due to forecasting user-specific trajectory information, rather than merely adding a soft prompt interface.

Problem

• personalized generation
• preference trajectory
• latent representation
• user drift
• frozen LLM

Innovation

• peer anchoring
• preference trajectory forecasting
• latent state extrapolation
• frozen LLM personalization
• soft token injection