// the take
A bibliography-style awesome-list tracking NeRF and 3D Gaussian Splatting papers applied to SLAM and robotics, spanning general NeRF/3DGS models through visual-SLAM, lidar-SLAM, multimodal-SLAM, and robotics tasks like manipulation and localization. It's for researchers trying to keep up with a fast-moving subfield, not for anyone looking for working code.
Genuinely exhaustive — entries run from the 2020 NeRF paper through 2026 venues (AAAI, ICRA, TITS), which means someone maintains it actively rather than letting it go stale. Good taxonomy: splitting NeRF-SLAM from 3DGS-SLAM, and both from general robotics use (manipulation/RL, planning, localization), makes it easier to find the right subsection than a flat chronological dump. Most entries link paper, code repo, project website, and video where they exist, so it functions as a real jumping-off point rather than just titles.
It's a single giant README with no search or filtering — at this entry count you're ctrl-F'ing a multi-thousand-line markdown file, which doesn't scale. Link quality is inconsistent: several entries have '[Code](todo)' placeholders or no code link at all, so you can't tell a reproducible method from a paper-only one without opening each PDF. There's exactly one actual piece of software in the whole repo (a single pip package under 'Tools') — everything else is a reading list, and the repo itself ships no scripts, benchmarks, or comparison tooling to act on the papers it indexes. Curation depends on PR volume from the community per the own contribution guide, so coverage and quality will vary by how much maintainer review each PR actually gets.
We read the README, the directory tree and a key source file in October 2026. Nobody pays to be listed. How we judge →