HyCell package README (Tencent Hunyuan, commit 0dcc533c)
GitHub repository Tencent-Hunyuan/Hyra-results at commit 0dcc533cc9f8f76c90ecbfe944df816b6672328a · Project source
What this source contributed
Records in this release that cite this source. A record can cite several sources.
Claims
Statements this source supports, each with where in the source it was checked.
- HyCell report: cell level scope: Scope The benchmark measures cell-level tasks; sample-level tasks such as predicting a donor's disease state, which Stack's use of the other cells of a sample helps, are not part of it. (Section 3.2 'Overall score'; README 'Recorded result')
- HyCell report: decoded expression: Outputs The decoder returns expected expression in counts per 10,000 over the genes it is given, using the code for an unknown dataset and donor. (README 'Decode')
- HyCell report: donor level properties: Scope HyCell embeds every cell on its own and is trained to discount donor and dataset effects; the report says properties shared by all cells of a donor, such as sex or disease state, are better read from embeddings aggregated per donor or sample. (Section 6; README 'Embed')
- HyCell report: embedding dimension: Outputs The embedding is a 512-dimensional linear projection of the final [CLS] state. (Section 2.2; Table 1 'Embedding')
- HyCell report: excluded atlases: Training evaluation overlap The seven atlases excluded from pretraining are the scIB human immune, pancreas and lung atlases (integer-count parts), PBMC from 10x Genomics pbmc4k and pbmc8k, the healthy donors of the MultiVINE brain vasculature atlas, a 40,322-cell sample of the cross-tissue immune atlas of Dominguez Conde et al. and a 41,140-cell sample of the Asian Immune Diversity Atlas. Every contributing dataset and collection was removed from the corpus. (Section 3.1 'Atlases'; Table 2; Section 2.5)
- HyCell report: finetuned predictions: Outputs A fine-tuned annotator returns cell-type predictions for a query. (README 'Fine-tune')
- HyCell report: input rules: Inputs Raw integer human counts; genes are matched by Ensembl id, otherwise by symbol, genes outside the vocabulary are dropped and duplicates summed. All measured genes must be passed, not a highly-variable subset, because measured zeros are drawn from the genes present in the input. (README 'Inputs')
- HyCell report: readme best s: Headline The README's results table marks Stack's S of 0.736 in bold as the best value of the column. (README 'Recorded result' table, column S)
- HyCell report: reproducibility: Reproducibility On a CUDA GPU with flash-attn the package reproduces the scored embeddings bit for bit, verified on all seven evaluation atlases (167,804 cells). (README 'Reproducibility')
- HyCell report: reproducibility cpu: Reproducibility Without flash-attn, PyTorch attention gives the same embeddings up to rounding, with per-cell cosine similarity above 0.9999. (README 'Reproducibility')
- HyCell report: throughput cpu: Throughput About 13 cells per second on CPU with 16 threads. (README 'Install')
- HyCell report: throughput gpu: Throughput About 1,500 cells per second on one NVIDIA H20 for atlases of 40,000 cells, including reading the file and tokenisation. Approximate and on unnamed atlases, so not recorded as a result. (README 'Recorded result', final paragraph)
Models, benchmarks and other records
Models (1)
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.
20 evidence rows matching the loaded filters
| Property and statement | Original source and location | Review and provenance |
|---|---|---|
| Archive note Archived as artifacts/hycell-README.md. Source metadata | HyCell package README (Tencent Hunyuan, commit 0dcc533c) No field-specific location recorded Version: GitHub repository Tencent-Hunyuan/Hyra-results at commit 0dcc533cc9f8f76c90ecbfe944df816b6672328a | catalogued No individual claim review recorded Audit detailsField: Source artifact SHA-256: Hash scope: Hash scope not separately documented; inspect source record |
| Artifact SHA-256 cd2eb7ae932124cacf9875b0ed1c48504b3e95b8e36169520bd1fe83cc0eecde Source metadata | HyCell package README (Tencent Hunyuan, commit 0dcc533c) No field-specific location recorded Version: GitHub repository Tencent-Hunyuan/Hyra-results at commit 0dcc533cc9f8f76c90ecbfe944df816b6672328a | catalogued No individual claim review recorded Audit detailsField: Source artifact SHA-256: Hash scope: Hash scope not separately documented; inspect source record |
| Artifact url https://raw.githubusercontent.com/Tencent-Hunyuan/Hyra-results/0dcc533cc9f8f76c90ecbfe944df816b6672328a/AI4Science/hycell/README.md Source metadata | HyCell package README (Tencent Hunyuan, commit 0dcc533c) No field-specific location recorded Version: GitHub repository Tencent-Hunyuan/Hyra-results at commit 0dcc533cc9f8f76c90ecbfe944df816b6672328a | catalogued No individual claim review recorded Audit detailsField: Source artifact SHA-256: Hash scope: Hash scope not separately documented; inspect source record |
| Licence Apache-2.0 (repository LICENSE) Source metadata | HyCell package README (Tencent Hunyuan, commit 0dcc533c) No field-specific location recorded Version: GitHub repository Tencent-Hunyuan/Hyra-results at commit 0dcc533cc9f8f76c90ecbfe944df816b6672328a | catalogued No individual claim review recorded Audit detailsField: Source artifact SHA-256: Hash scope: Hash scope not separately documented; inspect source record |
Limitations1 values
| HyCell package README (Tencent Hunyuan, commit 0dcc533c) No field-specific location recorded Version: GitHub repository Tencent-Hunyuan/Hyra-results at commit 0dcc533cc9f8f76c90ecbfe944df816b6672328a | catalogued No individual claim review recorded Audit detailsField: Source artifact SHA-256: Hash scope: Hash scope not separately documented; inspect source record |
| Media type text/markdown Source metadata | HyCell package README (Tencent Hunyuan, commit 0dcc533c) No field-specific location recorded Version: GitHub repository Tencent-Hunyuan/Hyra-results at commit 0dcc533cc9f8f76c90ecbfe944df816b6672328a | catalogued No individual claim review recorded Audit detailsField: Source artifact SHA-256: Hash scope: Hash scope not separately documented; inspect source record |
| Publication status official_project_source Source metadata | HyCell package README (Tencent Hunyuan, commit 0dcc533c) No field-specific location recorded Version: GitHub repository Tencent-Hunyuan/Hyra-results at commit 0dcc533cc9f8f76c90ecbfe944df816b6672328a | catalogued No individual claim review recorded Audit detailsField: Source artifact SHA-256: Hash scope: Hash scope not separately documented; inspect source record |
| Retrieved at 2026-11-06T11:42:11Z Source metadata | HyCell package README (Tencent Hunyuan, commit 0dcc533c) No field-specific location recorded Version: GitHub repository Tencent-Hunyuan/Hyra-results at commit 0dcc533cc9f8f76c90ecbfe944df816b6672328a | catalogued No individual claim review recorded Audit detailsField: Source artifact SHA-256: Hash scope: Hash scope not separately documented; inspect source record |
| Source type repository README Source metadata | HyCell package README (Tencent Hunyuan, commit 0dcc533c) No field-specific location recorded Version: GitHub repository Tencent-Hunyuan/Hyra-results at commit 0dcc533cc9f8f76c90ecbfe944df816b6672328a | catalogued No individual claim review recorded Audit detailsField: Source artifact SHA-256: Hash scope: Hash scope not separately documented; inspect source record |
| Url https://github.com/Tencent-Hunyuan/Hyra-results/blob/0dcc533cc9f8f76c90ecbfe944df816b6672328a/AI4Science/hycell/README.md Source metadata | HyCell package README (Tencent Hunyuan, commit 0dcc533c) No field-specific location recorded Version: GitHub repository Tencent-Hunyuan/Hyra-results at commit 0dcc533cc9f8f76c90ecbfe944df816b6672328a | catalogued No individual claim review recorded Audit detailsField: Source artifact SHA-256: Hash scope: Hash scope not separately documented; inspect source record |
Sources and history
Release 2026-11-21-adc15377fab2 · Record review: source checked
README of the hycell package, with the recorded benchmark result, method summary, installation, embedding, decoding, fine-tuning and input rules.
Technical metadata and extraction receipts
Stable ID: hycell-20261106-source-model-readme
- areas
- single-cell; cells-tissues
- contexts
- research
- archive note
- Archived as artifacts/hycell-README.md.
- artifact sha256
- cd2eb7ae932124cacf9875b0ed1c48504b3e95b8e36169520bd1fe83cc0eecde
- artifact url
- https://raw.githubusercontent.com/Tencent-Hunyuan/Hyra-results/0dcc533cc9f8f76c90ecbfe944df816b6672328a/AI4Science/hycell/README.md
- licence
- Apache-2.0 (repository LICENSE)
- limitations
- Its results table is a six-model subset of Table 3 and its fine-tuning table repeats Table 5; both are asserted equal to the report and not recorded again. Its throughput figures (about 1,500 cells per second on an H20, about 13 on CPU) are approximate and appear only here.
- media type
- text/markdown
- publication status
- official_project_source
- retrieved at
- 2026-11-06T11:42:11Z
- source type
- repository README
- url
- https://github.com/Tencent-Hunyuan/Hyra-results/blob/0dcc533cc9f8f76c90ecbfe944df816b6672328a/AI4Science/hycell/README.md
- version
- GitHub repository Tencent-Hunyuan/Hyra-results at commit 0dcc533cc9f8f76c90ecbfe944df816b6672328a