TRACER: Balancing Stability-Plasticity-Cognitivity Trilemma for LLM Enhanced Continual Recommendation
2026-08-17 • Information Retrieval
Information Retrieval
AI summaryⓘ
The authors study how recommendation systems can keep up with users' changing tastes over time, which is hard when data is sparse. They find a tricky balance called the Stability-Plasticity-Cognitivity Trilemma, where using knowledge from large language models (LLMs) can conflict with remembering past preferences and adapting to new interests. To solve this, they created TRACER, a system made of three parts that work together to handle these challenges without one overpowering the others. Tested on five datasets, TRACER improved recommendations compared to existing methods.
Continual RecommendationLarge Language Models (LLMs)Semantic KnowledgeStability-Plasticity-Cognitivity TrilemmaUser Interest EvolutionSparsityPersonalized PreferencesMachine Learning ModulesContinual Learning
Authors
WooJoo Kim, HyunSik Yoo, JunYoung Kim, JaeHyung Lim, SeongKu Kang, HwanJo Yu
Abstract
Continual recommendation aims to capture evolving user interests from streaming data but struggles with sparsity. LLM enhancers mitigate this with semantic knowledge, but naive integration creates a new conflict. We identify this as the Stability-Plasticity-Cognitivity (SPC) Trilemma, where generalized LLM semantic priors (Cognitivity) conflict with retaining personalized historical preferences (Stability) and adapting to individual interest shifts (Plasticity). To address this, we propose Trilemma-Responsive Adaptive Continual Enhancement for Recommendation (TRACER). TRACER synergistically combines three specialized modules, each targeting stability, plasticity, or cognitivity, while preventing any single lemma from dominating. This holistic design enables semantic knowledge to support history retention and adaptation to evolving interests without disrupting continual learning. Across five real-world datasets, TRACER effectively harmonizes the SPC trilemma and outperforms state-of-the-art baselines by up to 14.38%. Our code is available at https://github.com/woo-joo/TRACER_CIKM26.