{"schema_version":"onlylabs.public_signal.v1","title":"Microsoft Model: microsoft/scFM-dataselection","description":"Microsoft model signal with public source context, captured evidence pages, related signals, and data-business radar classification.","url":"https://onlylabs.fyi/signals/b3b6ebec-d08e-40ce-b4f8-d43a1ca65044","json_url":"https://onlylabs.fyi/signals/b3b6ebec-d08e-40ce-b4f8-d43a1ca65044/signal.json","generated_at":"2026-10-24T20:16:54.590Z","evidence_latest_fetched_at":"2026-10-17T07:20:24.410811+00:00","signal_first_seen_at":"2026-10-17T07:20:08.577778+00:00","org":{"slug":"microsoft","name":"Microsoft","category":"frontier-lab","category_label":"Frontier lab","dossier_url":"https://onlylabs.fyi/labs/microsoft","dossier_json_url":"https://onlylabs.fyi/labs/microsoft/dossier.json"},"related_urls":{"signal":"https://onlylabs.fyi/signals/b3b6ebec-d08e-40ce-b4f8-d43a1ca65044","signal_json":"https://onlylabs.fyi/signals/b3b6ebec-d08e-40ce-b4f8-d43a1ca65044/signal.json","source":"https://huggingface.co/microsoft/scFM-dataselection","lab_dossier":"https://onlylabs.fyi/labs/microsoft","lab_dossier_json":"https://onlylabs.fyi/labs/microsoft/dossier.json","analysis":"https://onlylabs.fyi/analysis/microsoft","analysis_json":"https://onlylabs.fyi/analysis/microsoft/analysis.json","analysis_evidence_json":"https://onlylabs.fyi/analysis/microsoft/evidence.json","category":"https://onlylabs.fyi/frontier","category_json":"https://onlylabs.fyi/frontier.json","category_feed":"https://onlylabs.fyi/frontier/feed.xml","category_signals_json":"https://onlylabs.fyi/signals.json","topic":"https://onlylabs.fyi/topics/releases","topic_signals_json":"https://onlylabs.fyi/topics/releases/signals.json","topic_feed":"https://onlylabs.fyi/topics/releases/feed.xml","data_business":{"radar":"https://onlylabs.fyi/data-radar","radar_json":"https://onlylabs.fyi/data-radar.json","opportunities":"https://onlylabs.fyi/opportunities","opportunities_json":"https://onlylabs.fyi/opportunities.json","lanes":[{"key":"data","label":"Data demand","url":"https://onlylabs.fyi/data-radar/data","json_url":"https://onlylabs.fyi/data-radar/data/signals.json"}]}},"answer_pack":{"answer":"Microsoft published microsoft/scFM-dataselection. This model signal is evidence of what shipped on model infrastructure and how the release is positioned. High-signal details: license mit · 0 HF downloads · Data selection tool for training single-cell foundation models.. onlylabs links this event to 1 captured evidence page and 6 related model signals. It also maps to Data demand in the data-business radar.","signal_desk":"releases","source_context":{"source_url":"https://huggingface.co/microsoft/scFM-dataselection","source_host":"huggingface.co","occurred_at":"2026-10-16T13:54:04+00:00","first_seen_at":"2026-10-17T07:20:08.577778+00:00","date_source":"source","context":null},"context_markers":[{"label":"Lab","value":"Microsoft","source":"signal"},{"label":"Signal desk","value":"releases","source":"signal"},{"label":"Source host","value":"huggingface.co","source":"source"},{"label":"License","value":"mit","source":"model"},{"label":"Downloads","value":"0","source":"traction"},{"label":"Likes","value":"2","source":"traction"},{"label":"Context","value":"Data selection tool for training single-cell foundation models.","source":"signal"},{"label":"Notability","value":"Microsoft releases niche single-cell data selection model.","source":"signal"},{"label":"Radar lane","value":"Data demand","source":"radar"},{"label":"Matched term","value":"data","source":"radar"},{"label":"Watch term","value":"Model card","source":"model"},{"label":"Watch term","value":"Data pipeline","source":"evidence"},{"label":"Watch term","value":"Infrastructure","source":"evidence"}],"evidence_coverage":{"target_pages":1,"captured_pages":1,"readable_pages":1,"capture_methods":["plain"],"missing_page_urls":[],"failed_page_urls":[],"blocked_page_urls":[],"page_urls":["https://huggingface.co/microsoft/scFM-dataselection/raw/main/README.md"],"related_signals":6,"has_source_url":true,"latest_page_fetched_at":"2026-10-17T07:20:24.410811+00:00"},"data_business":{"matches":true,"lanes":[{"key":"data","label":"Data demand","url":"https://onlylabs.fyi/data-radar/data","json_url":"https://onlylabs.fyi/data-radar/data/signals.json"}],"matched_terms":["data"],"score":13,"reason":"Microsoft has a model signal matching data demand."},"agent_handoff":{"signal_json":"https://onlylabs.fyi/signals/b3b6ebec-d08e-40ce-b4f8-d43a1ca65044/signal.json","dossier_json":"https://onlylabs.fyi/labs/microsoft/dossier.json","analysis_json":"https://onlylabs.fyi/analysis/microsoft/analysis.json","analysis_evidence_json":"https://onlylabs.fyi/analysis/microsoft/evidence.json","topic_signals_json":"https://onlylabs.fyi/topics/releases/signals.json","topic_feed":"https://onlylabs.fyi/topics/releases/feed.xml","category_signals_json":"https://onlylabs.fyi/signals.json","data_radar_json":"https://onlylabs.fyi/data-radar.json","opportunities_json":"https://onlylabs.fyi/opportunities.json"},"analysis_playbook":{"objective":"Turn model cards, repository releases, and launch artifacts into a current view of what shipped and what changed.","evidence_focus":["model card","release note","version or tag","pipeline tag","license","downloads, likes, stars, and linked papers"],"extraction_questions":["What actually shipped, and through which public artifact?","Which releases include inspectable model, repo, paper, or package evidence?","Which launches create new eval, data, infra, or deployment needs?","Which shipped signals are missing captured page context?"],"signal_questions":["What shipped in this model artifact, and what evidence describes its capability, license, or traction?","What actually shipped, and through which public artifact?","Which releases include inspectable model, repo, paper, or package evidence?","Which data-business lane explains this signal: Data demand?","Do the 6 related model signals show a repeated pattern?"],"output_fields":["org","release_type","artifact","traction","data_business_lane","evidence_url"],"data_business_relevance":"Release signals show the hard edge of lab execution and help connect data-business opportunities to concrete model, repo, and deployment changes.","required_sources":[{"label":"signal_json","url":"https://onlylabs.fyi/signals/b3b6ebec-d08e-40ce-b4f8-d43a1ca65044/signal.json","required":true},{"label":"source","url":"https://huggingface.co/microsoft/scFM-dataselection","required":true},{"label":"dossier_json","url":"https://onlylabs.fyi/labs/microsoft/dossier.json","required":true},{"label":"analysis_evidence_json","url":"https://onlylabs.fyi/analysis/microsoft/evidence.json","required":true},{"label":"topic_signals_json","url":"https://onlylabs.fyi/topics/releases/signals.json","required":false},{"label":"data_radar_json","url":"https://onlylabs.fyi/data-radar.json","required":true}],"expected_output":["one-paragraph source-grounded interpretation","data-business implication","confidence and missing evidence","recommended next source to inspect"],"prompt_seed":"Using only the linked onlylabs JSON, captured source context, and cited evidence, analyze Microsoft's model signal \"microsoft/scFM-dataselection\" for frontier lab strategy and data-business implications."},"semantic_triples":[{"subject":"Microsoft","predicate":"published model","object":"microsoft/scFM-dataselection","text":"Microsoft published model microsoft/scFM-dataselection."},{"subject":"microsoft/scFM-dataselection","predicate":"is classified as","object":"model signal","text":"microsoft/scFM-dataselection is classified as model signal."},{"subject":"microsoft/scFM-dataselection","predicate":"belongs to","object":"releases desk","text":"microsoft/scFM-dataselection belongs to releases desk."},{"subject":"microsoft/scFM-dataselection","predicate":"has evidence coverage","object":"1 captured evidence page","text":"microsoft/scFM-dataselection has evidence coverage 1 captured evidence page."},{"subject":"microsoft/scFM-dataselection","predicate":"matches data-business lanes","object":"Data demand","text":"microsoft/scFM-dataselection matches data-business lanes Data demand."},{"subject":"microsoft/scFM-dataselection","predicate":"has captured page count","object":"1","text":"microsoft/scFM-dataselection has captured page count 1."},{"subject":"microsoft/scFM-dataselection","predicate":"has readable page count","object":"1","text":"microsoft/scFM-dataselection has readable page count 1."},{"subject":"microsoft/scFM-dataselection","predicate":"has related signal count","object":"6","text":"microsoft/scFM-dataselection has related signal count 6."},{"subject":"microsoft/scFM-dataselection","predicate":"has analysis playbook objective","object":"Turn model cards, repository releases, and launch artifacts into a current view of what shipped and what changed.","text":"microsoft/scFM-dataselection has analysis playbook objective Turn model cards, repository releases, and launch artifacts into a current view of what shipped and what changed.."},{"subject":"microsoft/scFM-dataselection","predicate":"has source host","object":"huggingface.co","text":"microsoft/scFM-dataselection has source host huggingface.co."},{"subject":"microsoft/scFM-dataselection","predicate":"has lab","object":"Microsoft","text":"microsoft/scFM-dataselection has lab Microsoft."},{"subject":"microsoft/scFM-dataselection","predicate":"has signal desk","object":"releases","text":"microsoft/scFM-dataselection has signal desk releases."},{"subject":"microsoft/scFM-dataselection","predicate":"has source host","object":"huggingface.co","text":"microsoft/scFM-dataselection has source host huggingface.co."},{"subject":"microsoft/scFM-dataselection","predicate":"has license","object":"mit","text":"microsoft/scFM-dataselection has license mit."},{"subject":"microsoft/scFM-dataselection","predicate":"has downloads","object":"0","text":"microsoft/scFM-dataselection has downloads 0."},{"subject":"microsoft/scFM-dataselection","predicate":"has likes","object":"2","text":"microsoft/scFM-dataselection has likes 2."},{"subject":"microsoft/scFM-dataselection","predicate":"has context","object":"Data selection tool for training single-cell foundation models.","text":"microsoft/scFM-dataselection has context Data selection tool for training single-cell foundation models.."},{"subject":"microsoft/scFM-dataselection","predicate":"has notability","object":"Microsoft releases niche single-cell data selection model.","text":"microsoft/scFM-dataselection has notability Microsoft releases niche single-cell data selection model.."}]},"intelligence":{"signal_desk":"releases","answer":"Microsoft published microsoft/scFM-dataselection. This model signal is evidence of what shipped on model infrastructure and how the release is positioned. High-signal details: license mit · 0 HF downloads · Data selection tool for training single-cell foundation models.. onlylabs links this event to 1 captured evidence page and 6 related model signals. It also maps to Data demand in the data-business radar.","semantic_triples":[{"subject":"Microsoft","predicate":"published model","object":"microsoft/scFM-dataselection","text":"Microsoft published model microsoft/scFM-dataselection."},{"subject":"microsoft/scFM-dataselection","predicate":"is classified as","object":"model signal","text":"microsoft/scFM-dataselection is classified as model signal."},{"subject":"microsoft/scFM-dataselection","predicate":"belongs to","object":"releases desk","text":"microsoft/scFM-dataselection belongs to releases desk."},{"subject":"microsoft/scFM-dataselection","predicate":"has evidence coverage","object":"1 captured evidence page","text":"microsoft/scFM-dataselection has evidence coverage 1 captured evidence page."},{"subject":"microsoft/scFM-dataselection","predicate":"matches data-business lanes","object":"Data demand","text":"microsoft/scFM-dataselection matches data-business lanes Data demand."}]},"signal":{"id":"b3b6ebec-d08e-40ce-b4f8-d43a1ca65044","url":"https://onlylabs.fyi/signals/b3b6ebec-d08e-40ce-b4f8-d43a1ca65044","json_url":"https://onlylabs.fyi/signals/b3b6ebec-d08e-40ce-b4f8-d43a1ca65044/signal.json","source_url":"https://huggingface.co/microsoft/scFM-dataselection","title":"microsoft/scFM-dataselection","summary":"Microsoft published a model signal. onlylabs watches model cards, launch context, and traction.","context":null,"kind":{"key":"model_released","label":"Model"},"org":{"slug":"microsoft","name":"Microsoft","category":"frontier-lab"},"occurred_at":"2026-10-16T13:54:04+00:00","first_seen_at":"2026-10-17T07:20:08.577778+00:00","date_source":"source","evidence_coverage":{"target_pages":1,"captured_pages":1,"readable_pages":1,"capture_methods":["plain"],"missing_page_urls":[],"failed_page_urls":[],"blocked_page_urls":[],"page_urls":["https://huggingface.co/microsoft/scFM-dataselection/raw/main/README.md"]},"facets":{"license":"mit"},"traction":{"github_stars":null,"hn_points":null,"hn_comments":null,"hn_story_id":null,"hf_downloads":0,"hf_likes":2},"data_radar":{"lanes":[{"key":"data","label":"Data demand","url":"https://onlylabs.fyi/data-radar/data"}],"score":13,"matched_terms":["data"],"reason":"Microsoft has a model signal matching data demand."}},"primary_evidence_page":{"is_primary":true,"source_match":true,"url":"https://huggingface.co/microsoft/scFM-dataselection/raw/main/README.md","final_url":"https://huggingface.co/microsoft/scFM-dataselection/raw/main/README.md","title":"microsoft/scFM-dataselection model card","http_status":200,"content_type":"text/plain; charset=utf-8","capture_method":"plain","fetched_at":"2026-10-17T07:20:24.410811+00:00","bytes":3818,"raw_path":"1059ba7ebe6ef028522f548b33ac5e56d2c98fec7fabd8426922a2a2bd9c59b0.md","content_hash":"397fc88af0616b1020d43cc2b66c104b2d7822864ce85263acd6c2e982dc81c1","excerpt_chars":1200,"truncated":true,"excerpt":"scFM-dataselection pretraining checkpoints Model checkpoints and PCA components for **Evaluating the role of pretraining dataset size and diversity on single-cell foundation model performance**, DenAdel et al., *Nature Methods* (2026). - [Paper](https://doi.org/10.1038/s41592-026-03120-y) - [Code and preprocessing](https://github.com/microsoft/scFM-dataselection) Models | Model | Count | |---|---:| | Geneformer | 125 | | scVI | 125 | | SSL/MLP | 125 | | PCA components | 80 | | **Total** | **455** | The full download is approximately **22.4 GB**: 375 neural network checkpoints and 80 PCA component matrices. Within `portable-checkpoints/`, models are grouped into `geneformer/`, `scvi/`, `ssl/` and `pca/`. The Geneformer, scVI and SSL models cover random sampling, cell-type reweighting and geometric sketching at 1, 10, 25, 50, 75 and 100 percent of the pretraining corpus. The collection also includes perturbation spike-in experiments. See `model-index.csv` for model folders and training settings. Download Use Python **3.10**. The examples below were tested on CPU. ```python from huggingface_hub import snapshot_download root = snapshot_download( repo_id=\"microsoft/scFM-dataselection\",..."},"evidence_pages":[],"related_signals":[{"id":"3c5288c5-7c51-4bbd-b55f-aa0a46f41c6a","url":"https://onlylabs.fyi/signals/3c5288c5-7c51-4bbd-b55f-aa0a46f41c6a","source_url":"https://huggingface.co/microsoft/Sentiment-Analysis-7B","title":"microsoft/Sentiment-Analysis-7B","context":null,"kind":{"key":"model_released","label":"Model"},"org":{"slug":"microsoft","name":"Microsoft","category":"frontier-lab"},"occurred_at":"2026-10-17T00:51:18+00:00","first_seen_at":"2026-10-21T07:20:08.588594+00:00","date_source":"source"},{"id":"68d3ddcf-a3b5-4088-b910-839b0df3f140","url":"https://onlylabs.fyi/signals/68d3ddcf-a3b5-4088-b910-839b0df3f140","source_url":"https://huggingface.co/microsoft/UserRepGen-4B","title":"microsoft/UserRepGen-4B","context":null,"kind":{"key":"model_released","label":"Model"},"org":{"slug":"microsoft","name":"Microsoft","category":"frontier-lab"},"occurred_at":"2026-10-14T20:30:00+00:00","first_seen_at":"2026-10-15T07:20:08.604058+00:00","date_source":"source"},{"id":"135a1458-1286-4b05-8fed-6b192611df65","url":"https://onlylabs.fyi/signals/135a1458-1286-4b05-8fed-6b192611df65","source_url":"https://huggingface.co/microsoft/AesCode-32B","title":"microsoft/AesCode-32B","context":null,"kind":{"key":"model_released","label":"Model"},"org":{"slug":"microsoft","name":"Microsoft","category":"frontier-lab"},"occurred_at":"2026-09-29T13:05:36+00:00","first_seen_at":"2026-10-11T07:20:08.256841+00:00","date_source":"source"},{"id":"f18b0a2e-7e51-403a-b8f9-6374b4d0d817","url":"https://onlylabs.fyi/signals/f18b0a2e-7e51-403a-b8f9-6374b4d0d817","source_url":"https://huggingface.co/microsoft/AesCode-8B","title":"microsoft/AesCode-8B","context":null,"kind":{"key":"model_released","label":"Model"},"org":{"slug":"microsoft","name":"Microsoft","category":"frontier-lab"},"occurred_at":"2026-09-29T13:04:44+00:00","first_seen_at":"2026-10-11T07:20:08.256841+00:00","date_source":"source"},{"id":"d7c67358-0019-41e1-a212-b1ad62f572fe","url":"https://onlylabs.fyi/signals/d7c67358-0019-41e1-a212-b1ad62f572fe","source_url":"https://huggingface.co/microsoft/rho-roboeval","title":"microsoft/rho-roboeval","context":null,"kind":{"key":"model_released","label":"Model"},"org":{"slug":"microsoft","name":"Microsoft","category":"frontier-lab"},"occurred_at":"2026-09-21T19:13:53+00:00","first_seen_at":"2026-09-23T07:20:08.566613+00:00","date_source":"source"},{"id":"add52401-a2c9-4f24-ae17-48bb89c2cae6","url":"https://onlylabs.fyi/signals/add52401-a2c9-4f24-ae17-48bb89c2cae6","source_url":"https://huggingface.co/microsoft/FrogNano-4B-2609","title":"microsoft/FrogNano-4B-2609","context":null,"kind":{"key":"model_released","label":"Model"},"org":{"slug":"microsoft","name":"Microsoft","category":"frontier-lab"},"occurred_at":"2026-09-17T00:13:57+00:00","first_seen_at":"2026-10-03T07:20:08.155123+00:00","date_source":"source"}]}