Evaluating Factual Consistency in Long-Document Summaries
Explore challenges and improvements in factual consistency metrics for long-document summarization using advanced perturbation testing methods.
EvoDev: Iterative Feature-Driven Software Dev with LLM Agents
Discover EvoDev, an iterative, feature-driven framework using LLM agents to automate end-to-end software development with improved efficiency and agility.
FedPF: Balancing Privacy, Fairness & Utility in Federated Learning
Discover FedPF, a federated learning algorithm that ensures privacy and fairness while maintaining high accuracy and low bias in AI models.
Solving Entropy Collapse in RLVR with STEER Method
Discover how STEER tackles entropy collapse in RLVR, enhancing LLM training by improving entropy dynamics and boosting reasoning performance.
Process Reward Models for Large Language Models Survey
Explore Process Reward Models in LLMs, from outcome signals to process supervision, enhancing reasoning and alignment in AI systems.