Existing steering benchmarks are built around simplified, isolated generation tasks and evaluate a single behavioral attribute under controlled settings that do not reflect the diversity and complexity of practical deployment. As a result, performance on these benchmarks provides only a limited picture of how steering...
Xin-Yi Cai, Bin Hu, Ming-Yu Jin et al.· 0 citations
Evaluations of fifteen LLM doctor agents show that moving from a single-turn diagnosis on the standardized records to multi-turn consultations causes large degradations, highlighting equity risks that single-turn benchmarks miss.
Youssef Mohamed, Ahmed Heakl, Qin-Rong Cui et al.· 0 citations
The Graph Theory Agent (GTA), which pairs a preference-trained representation selector with plan-and-decompose scaffolding around a frozen executor LLM, is proposed, which lifts Phi-4 from 53.5% to 69.1% on the benchmark's easy split and from 33.0% to 41.5% on its hard split.
Zi-Xiang Xu, Yan-Bo Wang, Chenxi Wang et al.· 4 citations· ⚡1
SocialMaze is introduced, a benchmark that organizes six tasks across social deduction games, daily-life interactions, and digital community platforms along three descriptive design axes: deep reasoning, dynamic interaction, and information uncertainty.
Zi-Xiang Xu, Yan-Bo Wang, Yue Huang et al.· 1 citation
We use cookies to run the site and, with your consent, for analytics and to show ads.
See our Cookie Policy.
