SOURCE-LINKED INTELLIGENCE

Drift-Constrained Optimization: Only Direction Matters in Fine-Tuning Instruct Models

arXiv · AI, language, vision and robotics · article · Sep 12, 2026 · UTC

Fine-tuning instruct models often improves target performance while inducing behavioral drift from the reference model, which can degrade existing capabilities. Rather than treating this drift as an uncontrolled consequence of optimization, we specify a behavioral drift budget before optimization and ask how to boost the target-task performance within it. Locally, behavioral drift induces a shared geometry anchored at the reference model, with the drift budget defining a boundary within this space. In this space, drift determines distance from the reference, leaving update direction as the rem

Read original source ↗ Open in workspace

recordType

paper

region

Global