arXiv:2611.11695 · v1
Language model capabilities fail in a shared, predictable order under compression and perturbation
摘要
Compression can preserve a language model's average accuracy while selectively degrading individual capabilities. Whether these losses follow method-specific patterns or a transferable order is unclear. We applied seven perturbation and compression methods to eight instruction-tuned models, scoring 16 tasks in four answer formats against preregistered predictions. Short-answer capabilities failed in nearly the same order under four random perturbations with different mechanisms (median correlation 0.90) and, within each model, under pruning and quantization (0.875). Intact answer margin, answer frequency and answer format predicted this order. Fragility profiles measured under random perturbation predicted which tasks and facts compression removes first and, after four-task calibration, held-out task accuracy (median absolute error 0.11-0.12, against 0.20-0.21 for uniform decline); profiles borrowed from other model families did almost as well. The order did not transfer across models under GPTQ quantization or hold for long-form tasks. Within short answers, what compression removes is predictable before deployment.
作者备注：45 pages (34 main text with Methods and Extended Data, 11 Supplementary Information), 4 figures, 6 Extended Data figures. Code, protocols and data: https://github.com/ZhixiangWang-CN/damage-order-artifacts ; interactive summary: https://zhixiangwang-cn.github.io/damage-order-artifacts/