Production coding prompts stack multiple constraints (output format, persona, urgency) at once. This shows those constraints interact non-additively and degrade reliability in model-specific ways, meaning testing constraints individually can miss real failure modes.

• articleDo Influence Tactics Matter? Investigating Prompt Framing Effects in LLM Code GenerationAlex Deaconu, Anubhav Gupta, Manaal Basha, Nicholas Haydu, Gema Rodr\'iguez-P\'erez
• articleBeyond Prompt Engineering: A Systematic Analysis of Prompt Lexical Sensitivity and Its Impacts on QualityQipeng Xie, Zi Liang, Jiafei Wu, Yufei Chen, Weizheng Wang, Wenao Ma, Zhong Ming, Haiqin Yang, Kaishun Wu
• articleHow Do Prompt Variations Affect Energy Consumption in On-Device LLMs?Wei Hu, Xiaolong Tu, Dawei Chen, Yitao Chen, Kyungtae Han, Haoxin Wang

Checking sign-in…

Loading comments…