Sparse Voxels Rasterization: Real-Time High-Fidelity Radiance Field Rendering(svraster.github.io)

103 points by jasondavies 1 year ago | 18 comments

ReconFusion: 3D Reconstruction with Diffusion Priors(reconfusion.github.io)

98 points by GaggiX 2 years ago | 13 comments

Lumiere: A space-time diffusion model for realistic video generation(lumiere-video.github.io)

282 points by jonbaer 2 years ago | 116 comments

InstantSplat: Sparse-View SfM-Free Gaussian Splatting in Seconds(instantsplat.github.io)

85 points by smusamashah 2 years ago | 2 comments

Sequential modeling enables scalable learning for large vision models(yutongbai.com)

126 points by og_kalu 2 years ago | 51 comments

Triangle splatting: radiance fields represented by triangles(trianglesplatting.github.io)

147 points by ath92 1 year ago | 64 comments

RealFill: Image completion using diffusion models(realfill.github.io)

557 points by flavoredquark 3 years ago | 178 comments

Project page for RealFill, a Google Research and Cornell paper on reference-driven image completion using diffusion models.

OpenWAM: An Open Framework for Composable World-Action Models(openwam.stanford.edu)

9 points by ilreb 19 days ago | 3 comments

Project page for OpenWAM, a Stanford open framework combining video prediction and robot control via composable world-action models on a causal video backbone.

Training language models to be warm and empathetic makes them less reliable(arxiv.org)

334 points by Cynddl 1 year ago | 333 comments

arXiv paper finding that fine-tuning language models to be warm and empathetic raises error rates and makes them more sycophantic on safety-critical tasks.

MeshGPT: Generating triangle meshes with decoder-only transformers(nihalsid.github.io)

684 points by jackcook 2 years ago | 148 comments

Project page for MeshGPT, a research paper from TU Munich, Politecnico di Torino and Audi that generates triangle meshes using decoder-only transformers.

Stealing Reasoning Traces from Proprietary LLM APIs(stolen-thoughts.com)

665 points by quantumgarbage 2 months ago | 295 comments

Project page for a research paper on extracting hidden reasoning traces from proprietary LLM APIs, by authors from MATS, ELLIS, Max Planck and Tübingen.

TurboQuant: Redefining AI efficiency with extreme compression(research.google)

495 points by ray__ 7 months ago | 134 comments

Google Research blog post introducing TurboQuant, a method for extreme compression to improve AI model efficiency.