Bootstrap stability of LoRA gains is not uniform: the ROUGE-1 confidence interval crossed zero at 7B-r4, so that increase is a trend rather than a stable improvement¶

Claim · All claims
Evidence · 1 study · causal r2 · q2
Currency · newest study 2026 · not yet checked against its sources · not reviewed by a person · recheck by 2029: tool and model accuracy changes quickly

Subclaims¶

q2 i? The paired bootstrap 95% confidence interval for ROUGE-1 improvement crossed zero at 7B-r4, unlike most other settings. → Shaowen Lu 2026

Evidence¶

Shaowen Lu 2026¶

Shaowen Lu, Chengxu Liu, Ping Zhou and Tao Yang. (2026). LoRA Fine-Tuned Models for Control Systems Course Q&A: A Multidimensional Evaluation of Model Scale and Rank Effects. https://arxiv.org/abs/2609.13918 doi:10.48550/arxiv.2609.13918

q2 · i? · causal · r2

Paired bootstrap robustness analysis (B=10,000 resamples of the 54-item test set; Fig. 4). The article reports the ROUGE-1 interval crossed zero at 7B-r4 while 7B intervals for all three ROUGE metrics were above zero at r=8 and r=16.

"In contrast, the confidence interval for ∆ROUGE-1 crossed zero at 7B-r4. The ROUGE-1 increase at that setting should therefore be interpreted as a trend and was less stable than the improvements under the other settings."

Discussion¶

Related Claims¶

• LoRA fine-tuning on control systems course data increases lexical and semantic similarity between model responses and course reference answers at both 3B and 7B model sizes — related
• Base-model size effects persist after LoRA adaptation: the 7B model gained more on ROUGE-2 and ROUGE-L, yet 3B-r16 outperformed 7B-r4 on ROUGE-L and BERTScore-F1 — related
• Increasing LoRA rank raises absolute ROUGE-L gains but produces diminishing returns in performance per million adapter parameters — related
• LoRA fine-tuning makes the Solution-Method-Teaching Points instructional format more stable, with Solution and Method coverage reaching 1.00 at r=8 and r=16 for the 3B model — related