AI summaryⓘ
The authors studied how personalization in large language models (LLMs) not only changes what the model says but also how it reasons when answering open-ended questions. They created a method called DRIFTLENS to measure how user-specific memory affects the reasoning steps compared to responses without that memory. Their experiments showed that incorporating user attributes like age or occupation causes noticeable changes in reasoning, even if the final answers still seem good. They also tested ways to reduce this reasoning drift after training, finding mixed results depending on the method and model. Overall, they highlight that personalized memory can shift reasoning in measurable ways and is not fully resolved by current techniques.
large language modelspersonalizationreasoning trajectoryuser attributescontext injectionopen-ended questionsDRIFTLENSreasoning driftpost-training methodsinstruction following
Authors
Xi Fang, Weijie Xu, Yingqiang Ge, Yuhui Xu, Stephanie Eckman, Chandan K. Reddy
Abstract
Personalization changes what a model says to a user; we show that it can also change the reasoning trajectory used to justify the response. Modern LLMs personalize interactions by storing user attributes, preferences, and prior context, then injecting this information into future prompts. We study whether such memory reshapes reasoning on open-ended questions where no single ground-truth answer exists. To quantify this effect, we introduce DRIFTLENS, a ground-truth-free framework that maps each expressed reasoning step to a value category and measures divergence between a question's no-memory trajectory and its trajectory under injected user-attribute memory. We first validate that DRIFTLENS distinguishes content-free pragmatic noise from substantive reasoning changes. Across four LLMs and 10 user-attribute categories, including age, occupation, and disability, user-attribute memory induces medium-to-large reasoning drift above each model's pragmatic-noise floor, even when final answers remain fluent, on-topic, and plausible. We then evaluate GRPO- and DPO-based post-training methods for reducing drift. Both reduce drift, but neither uniformly dominates; effects on downstream capability, helpfulness, and instruction following are model-and reward-dependent. These results suggest that memory-induced reasoning drift is a measurable and only partly mitigated failure mode of personalized language models.