Showing cs.SEShow all
3 papers · 1 filter
cs.SE2026
Route Before Response: Fail-Closed Routing for LLM-Enabled Customer-Service Software
Yuchen Cao, Zhenyu Mao, Jacky Wai Keung +3
LLM-enabled customer-service software in financial institutions must choose a downstream service before responding. Classifiers and LLM routers can identify a relevant label or des…
cs.SE2026
What Was Once Learned May Need to Be Unlearned: Machine Unlearning for Deprecated API Knowledge in Large Language Models
Jin Liu, Yanzhong He, Guancheng Lin +4
Large language models (LLMs) for code completion may generate deprecated APIs because their pre-training corpora contain code from historical library versions. Existing approaches…
cs.SE2026
Keep Evaluation Fair: Detecting Data Leakage in Code Generation Benchmarks via Membership Inference Attacks
Dongdong Zhao, Jian Chen, Guancheng Lin +3
Code generation benchmarks are widely used to evaluate Large Language Models (LLMs), but benchmark data leakage into training sets can inflate performance and undermine evaluation…