Africa
Inference-Time Intervention: Eliciting Truthful Answers from a Language Model Kenneth Li
This intervention significantly improves the performance of LLaMA models on the TruthfulQA benchmark.
ODGS: 3D Scene Reconstruction from Omnidirectional Images with 3D Gaussian Splatting
Omnidirectional (or 360-degree) images are increasingly being used for 3D applications since they allow the rendering of an entire scene with a single image.
LISA: Layerwise Importance Sampling for Memory-Efficient Large Language Model Fine-Tuning
Y et, their massive memory consumption has become a significant roadblock to large-scale training.
RDumb: A simple approach that questions our progress in continual test-time adaptation
Ori Press, Steffen Schneider, Matthias Kümmerer, Matthias Bethge
In addition, we introduce a simple baseline, "RDumb", that periodically resets the
Discrete Object Generation with Reversible Inductive Construction
Ari Seff, Wenda Zhou, Farhan Damani, Abigail Doyle, Ryan P. Adams
Neural Information Processing Systems http://nips.cc/