AI & ML interests

None defined yet.

Recent Activity

hletrd  updated a model about 2 hours ago
Xylolabs/gemma-4-E2B-it-GGUF
hletrd  updated a model about 2 hours ago
Xylolabs/gemma-4-E4B-it-GGUF
hletrd  updated a model about 2 hours ago
Xylolabs/NVIDIA-Nemotron-3.5-Lightning-30B-A3B-GGUF
View all activity

Organization Card
Xylolabs

Xylolabs Inc.

Edge Physical AI that catches equipment faults and runs robots.

Sensors, edge AI and robots for port cranes, power plants and factories.

Website GitHub Request a PoC

On Hugging Face

llama.cpp GGUF quantizations of popular open models, including Hy4, Hy4-flash, Inkling, Qwen3.8, Gemma 4, KORMo, clairic and Aion. Each card gives the KL divergence and top-1 agreement against the unquantized model, and charts of accuracy by file size against other quantizers.

MLX builds of some of them, for Apple silicon.

Tenstorrent Tensix builds of language, speech, image and robot models, in the block floating point formats bfloat4_b (BFP4) and bfloat8_b (BFP8) that TT-NN packs without changing the weights. Each card lists the format of every tensor class.

The Tensix code is on GitHub, Apache-2.0: tensix-bfp-gptq quantizes and checks the weights; tensix-kernels and tensix-runner hold the kernels and the runner for Tensix. Releases: tensix-bfp-gptq v0.1.0, tensix-kernels v0.1.0, tensix-runner v0.1.0.

Collections: one per model family.

What we build

Platform What it does
XyloZero Edge AI that reads sensor signals with PLC and drive data to catch equipment anomalies early
Battery Fire Early Detection Hears the acoustic precursors of thermal runaway in AGV, AMR and ESS batteries
Autonomy & Robotics On-device AI for robot and drone patrols, VLA manipulation learning and port crane autonomy

Started as signal research at Seoul National University in 2019, incorporated in December 2025.

datasets 0

None public yet