作者位次 #2/2ICML 2026Poster
Position: Sycophancy is an Educational Safety Risk: Why LLM Tutors Need Sycophancy Benchmarks
Enkelejda Kasneci · Gjergji Kasneci
首次获取该候选人的完整档案及 StartLux 评估共 20 积分；每位用户单独解锁，之后查看免费。主动刷新 20 积分。
实时查找电子邮箱，费用不超过20积分；未找到邮箱不扣费。
Enkelejda Kasneci · Gjergji Kasneci
Zheyu Zhang · Shuo Yang · Bardh Prenkaj · Gjergji Kasneci
Shuo Yang · Zheyu Zhang · Bardh Prenkaj · Gjergji Kasneci
Generating high-fidelity synthetic tabular data remains a critical challenge for enhancing data availability in privacy-sensitive and low-resource domains. Recent approaches leverage LLMs by representing table rows as sequences, yet suffer from two fundamental limitations: (1) they model feature...
Chenchen Yuan · Zheyu Zhang · Gjergji Kasneci
Large language models often display heterogeneous moral preferences across settings. We study inference-time steering toward a desired ethical framework while preserving general competence. We present Convergent-Divergent Routing, which traces and edits minimal branch points inside transformer bl...
Yuxiao Li · Alina Fastowski · Efstratios Zaradoukas · Bardh Prenkaj · Gjergji Kasneci
Activation steering has emerged as a powerful tool to shape LLM behaviour without the need for weight updates. While its inherent brittleness and unreliability are well-documented, its safety implications remain underexplored. In this work, we present a systematic safety audit of steering vectors...