AIBearisharXiv – CS AI · 3h ago7/10
🧠
Models Recall What They Violate: Constraint Adherence in Multi-Turn LLM Ideation
Researchers introduce DriftBench, a benchmark evaluating how well large language models maintain fidelity to original constraints during multi-turn iterative refinement. The study reveals a critical disconnect: models can accurately restate constraints while simultaneously violating them, with non-compliance rates ranging from 8% to 99% depending on the model.