HyCell
HyCell maps the raw counts of a human cell to a 512-dimensional embedding with a 16-million-parameter set transformer and decodes embeddings back to expected expression.
Results are recorded on linked configurations and versions of this model. Their settings and evaluations are kept separate below.
Results on linked records
This model has no results of its own in this release. 2 linked records have results. They are not assigned to the underlying model: each record keeps its own settings and evaluations, and its scores are on its own page.
- HyCell (HyCell report) Configuration · configuration of this model · 38 evaluations · 38 results
- HyCell fine-tuned annotator (HyCell report) Configuration · configuration of this model · 4 evaluations · 4 results
Overview
An explanatory profile has not yet been reviewed for this record.
Consult the linked sources for architecture or protocol details. Missing evidence is not evidence of a missing capability.
HyCell has no ranked result on a use-case protocol in this release.
Recorded parameters: Not recorded. Licence: Apache-2.0 for the code and weights. The hycell NOTICE opens "HyCell / Copyright (C) 2026 Tencent. Licensed under the Apache License, Version 2.0; see the LICENSE file at the root of this repository", the package README's License section says Apache-2.0 and states that the weights are part of the package, and CITATION.cff and pyproject.toml give Apache-2.0; the repository LICENSE's default for Tencent-authored materials backs all of them. No file names the weights file itself, and the package ships no LICENSE of its own. The training data (CZ CELLxGENE Census, release of 8 November 2025) and the gene vocabulary drawn from it are CC BY 4.0 and keep their own terms..
Model Capabilities & Hardware Requirements
Deployment facts recorded for HyCell. Fields the record does not carry are marked not recorded.
Evaluated Checkpoint: hycell-16m, package version 0.1.0, released 2026-11-02; weights SHA-256 11197d805aa6bafd7a69996bf8e0c7cfd7706b4fbbf5466f9e350d5b3e79d744; technical report dated 2026-11-01
Use this model
How it works, versions and access
Additional methodological details, versions and access terms have not yet been extracted into a structured profile for this model. Consult the linked sources and evaluation records below.
Evidence
Source checking verifies the cited claim or transcription. It does not establish independent reproduction.
Evidence table
Inspect claims, sources and review details
Trace each statement to its source and review. A context-only reference supports the record generally; it does not verify an individual field. Source checking does not reproduce an experiment.
One row per statement and cited source. Multiple citations are not independent evaluations. Shared locators are labelled explicitly.
0 evidence rows matching the loaded filters
| Property and statement | Original source and location | Review and provenance |
|---|
No evidence rows match these filters. Choose another scope or clear the search.
Sources and history
Release 2026-11-21-c8bea25e5abb · Record review: source checked
9 source records and release history
- HyCell: a Compact Embedding Model for Single-Cell Transcriptomes (Hyra Team, 2026-11-01) · Original source · Dated 2026-11-01 on page 1; retrieved at commit 0dcc533cc9f8f76c90ecbfe944df816b6672328a (9 pages)
- HyCell package README (Tencent Hunyuan, commit 0dcc533c) · Original source · GitHub repository Tencent-Hunyuan/Hyra-results at commit 0dcc533cc9f8f76c90ecbfe944df816b6672328a
- Hyra Results repository LICENSE (commit 0dcc533c) · Original source · GitHub repository Tencent-Hunyuan/Hyra-results at commit 0dcc533cc9f8f76c90ecbfe944df816b6672328a
- HyCell NOTICE (commit 0dcc533c) · Original source · GitHub repository Tencent-Hunyuan/Hyra-results at commit 0dcc533cc9f8f76c90ecbfe944df816b6672328a
- HyCell CITATION.cff (commit 0dcc533c) · Original source · GitHub repository Tencent-Hunyuan/Hyra-results at commit 0dcc533cc9f8f76c90ecbfe944df816b6672328a
- HyCell pyproject.toml (commit 0dcc533c) · Original source · GitHub repository Tencent-Hunyuan/Hyra-results at commit 0dcc533cc9f8f76c90ecbfe944df816b6672328a
- HyCell checkpoint sidecar weights.json (commit 0dcc533c) · Original source · GitHub repository Tencent-Hunyuan/Hyra-results at commit 0dcc533cc9f8f76c90ecbfe944df816b6672328a
- HyCell weights hycell-16m weights.safetensors (commit 0dcc533c) · Original source · GitHub repository Tencent-Hunyuan/Hyra-results at commit 0dcc533cc9f8f76c90ecbfe944df816b6672328a
- Hyra Results repository README (Tencent Hunyuan, commit 0dcc533c) · Original source · GitHub repository Tencent-Hunyuan/Hyra-results at commit 0dcc533cc9f8f76c90ecbfe944df816b6672328a
Technical metadata and record identifiers
Technical metadata for this record is tracked under stable ID hycell-20261106-model-hycell.
Source locator: Technical report Sections 2 and 2.5, Table 1; README 'Install', 'License'; weights.json; NOTICE
Related records
- subject: HyCell report: cell presentations
- subject: HyCell report: census release
- subject: HyCell report: corpus cells
- subject: HyCell report: dataset code entries
- subject: HyCell report: decoded expression
- subject: HyCell report: developer
- subject: HyCell report: development set
- subject: HyCell report: donor code entries
- subject: HyCell report: donor level properties
- subject: HyCell report: embedding dimension
- subject: HyCell report: expressed genes
- subject: HyCell report: finetune time
- subject: HyCell report: finetuned predictions
- subject: HyCell report: gene table parameters
- subject: HyCell report: heads
- subject: HyCell report: input rules
- subject: HyCell report: labels in training
- subject: HyCell report: layers
- subject: HyCell report: licence code
- subject: HyCell report: licence data
- subject: HyCell report: max tokens
- subject: HyCell report: measured zeros
- subject: HyCell report: parameters
- subject: HyCell report: protein coding
- subject: HyCell report: release date
- subject: HyCell report: reproducibility
- subject: HyCell report: reproducibility cpu
- subject: HyCell report: set encoding
- subject: HyCell report: swiglu width
- subject: HyCell report: throughput cpu
- subject: HyCell report: throughput gpu
- subject: HyCell report: training cells
- subject: HyCell report: training steps
- subject: HyCell report: training time
- subject: HyCell report: training tokens
- subject: HyCell report: transformer parameters
- subject: HyCell report: validation cells
- subject: HyCell report: vocabulary
- subject: HyCell report: width
- configuration of: HyCell (HyCell report)
- configuration of: HyCell fine-tuned annotator (HyCell report)
How to cite
HyCell. Rewire biological model benchmark database, release 2026-11-21-c8bea25e5abb. https://benchmarks.rewirebio.io/database/model/hycell-20261106-model-hycell/
Release 2026-11-21-c8bea25e5abb. Rankings and values change between releases, so keep the release id in the citation. Cite the original publication as well for any value you reuse: each result links its source.
The compilation and annotations made by Rewire in this release are offered under CC BY 4.0; values and text taken from cited sources stay under the terms of those sources and are not covered by this licence. Data licence