Infusion: Influence Functions to Edit AI Training Data
Discover how Infusion uses influence functions to edit training data, shaping AI model behavior for improved performance and security.
EB-JEPA: Lightweight Energy-Based Joint-Embedding Library
Discover EB-JEPA, a lightweight library for energy-based joint-embedding predictive architectures enabling efficient self-supervised learning and video mod...
SPICE: Efficient Data Selection for Large Language Model Training
SPICE improves large language model training by selecting conflict-aware data subsets, boosting performance while reducing training costs significantly.
Advanced Speech Editing Detection with Audio LLMs
Discover how prior-enhanced Audio LLMs unify speech editing detection and content localization for improved accuracy and robustness.
How Computer Environments Boost Agentic Intelligence in LLMs
Discover how computer environments enhance agentic intelligence and efficiency in large language models, improving performance across diverse AI tasks.