Constarium
← Search

Data · dataset · 2026

Xyntetik-Kvist-14B training record: width pruning, distillation and the envelope gate

Listed in Hugging Face Datasets

Xyntetik-Kvist-14B: training record, gate measurements and code Research evidence dataset.

Description

No model weights. Part of the collection Xyntetik Kvist: distilled agents on this account, produced with Xyntetik Runner.

Dataset summary Question tested. Whether Muse-Glimmer-30B's language model, cut by width to a dense 14.4 B-parameter student and distilled back from its frozen BF16 parent, keeps a working model (K1), recovers fast enough to be worth finishing (K2) and… See the full description on the dataset page: huggingface.co/datasets/Joakimpalm-Zen/Xyntetik-Kvist-14B-training-record.

Links

Topics

Provenance · 1 source records, 8 field assertions
SourceKeyLast seenRaw
Hugging Face DatasetsJoakimpalm-Zen/Xyntetik-Kvist-14B-training-record11 d agoJSON v1
FieldAssertionExtractorEvidence
access_levelsource · Hugging Faceconnector:huggingface@1.0.0/gated
concepts[field].local:field:computer-science-aimapping · Hugging Faceconnector:huggingface@1.0.0
created_datesource · Hugging Faceconnector:huggingface@1.0.0
descriptionsource · Hugging Faceconnector:huggingface@1.0.0/description
licensesource · Hugging Faceconnector:huggingface@1.0.0/tags[license:*]
publication_datesource · Hugging Faceconnector:huggingface@1.0.0
titlesource · Hugging Faceconnector:huggingface@1.0.0/id
updated_datesource · Hugging Faceconnector:huggingface@1.0.0