Benchmarking Educational LLMs for Gender Bias in Feedback
Explore how educational LLMs exhibit gender bias in feedback and learn methods to benchmark and ensure fairness in AI-driven teaching tools.
E-Scores: Robust Correctness Assessment for AI Outputs
Discover how e-scores improve the evaluation of generative AI outputs, ensuring flexible, reliable correctness assessment beyond traditional methods.
Fair Payoffs in Indivisible Games Using Shapley Value
Discover how the indivisible Shapley value ensures fair payoffs in coalitional games across politics, healthcare, and AI applications.
BIOGEN: Multi-Agent Framework for AMR Transcriptomic Analysis
Discover BIOGEN, a multi-agent reasoning framework offering evidence-based transcriptomic interpretation to advance antimicrobial resistance research.
Reducing Incoherence in Goal-Conditioned Autoregressive Models
Explore how re-training reduces incoherence in goal-conditioned autoregressive models, improving reinforcement learning policy performance and returns.