SOURCE-LINKED INTELLIGENCE

LOCUS: Task-Aware Low-Rank Post-Training for Token-Efficient Language Generation

arXiv · AI, language, vision and robotics · article · Sep 10, 2026 · UTC

Large language model serving costs scale directly with output sequence length, yet standard preference alignment often inflates response verbosity without improving utility. We study whether the parameterization of post-training updates affects generation length: low-rank subspaces alter sequence length without modifying the alignment loss. We present LOCUS, a method that selects a task-aware low-rank adaptation subspace to minimize output-token cost subject to a utility constraint. Within this subspace, post-training retains the native preference objective with a frozen backbone. Across Anthr

Read original source ↗ Open in workspace

recordType

paper

region

Global