Interpretability for LM post-training at three altitudes: attention, input attribution, and feature circuits. One diff-two-checkpoints verb.

A Claude Code channel that lets Claude sessions message each other across devices on the same local network.

An Open Claw inspired autonomous agent system built on top of Claude Code.

[CVPR2023] The official repo for OC-SORT: Observation-Centric SORT on video Multi-Object Tracking. OC-SORT is simple, online and robust to occlusion/non-linear motion.

Implementation of paper - YOLOv7: Trainable bag-of-freebies sets new state-of-the-art for real-time object detectors

Official PyTorch implementation of "Camera Distance-aware Top-down Approach for 3D Multi-person Pose Estimation from a Single RGB Image", ICCV 2019

Ultralytics YOLOv5 in PyTorch > ONNX > CoreML > TFLite

BoxMOT: Pluggable python and c++ SOTA multi-object tracking modules with support for axis-aligned and oriented bounding boxes

This repo houses the showcase for progress updates of the roundnet-ai private repo.

KAPAO is an efficient single-stage human pose estimation model that detects keypoints and poses as objects and fuses the detections to predict human poses.