REVIEW 2 cited by
LoRe: Personalizing LLMs via Low-Rank Reward Modeling
Not yet reviewed by Pith; the record is open.
This paper has not been read by Pith yet. Machine review is queued; the pith claim, tier, and objections will appear here once it completes.
SPECIMEN: schema-true, not a live event
T0 review · schema-true
One-sentence machine reading of the paper's core claim.
pith:XXXXXXXX · record.json · timestamp
read the original abstract
Personalizing large language models (LLMs) to accommodate diverse user preferences is essential for enhancing alignment and user satisfaction. Traditional reinforcement learning from human feedback (RLHF) approaches often rely on monolithic value representations, limiting their ability to adapt to individual preferences. We introduce a novel framework that leverages low-rank preference modeling to efficiently learn and generalize user-specific reward functions. By representing reward functions in a low-dimensional subspace and modeling individual preferences as weighted combinations of shared basis functions, our approach avoids rigid user categorization while enabling scalability and few-shot adaptation. We validate our method on multiple preference datasets, demonstrating superior generalization to unseen users and improved accuracy in preference prediction tasks.
Forward citations
Cited by 2 Pith papers
-
Cautious Context Steering for Language Model Personalization
CCS is a learned per-token gate for context steering that improves personalized generation on PRISM and four out-of-distribution benchmarks while avoiding a second forward pass per decoding step.
-
A Personalized Conversational Benchmark: Towards Simulating Personalized Conversations
PERSONACONVBENCH is a new Reddit-based benchmark showing that LLMs predict sentiment, community scores, and next replies better when given a user's multi-turn conversation history, and it releases public data and code.
Discussion (0). Continue with ORCID to comment.