{"schema_version":"onlylabs.public_signal.v1","title":"AI21 Labs Repo: AI21Labs/pmi-masking","description":"AI21 Labs repo signal with public source context, captured evidence pages, related signals, and category-scoped analysis context.","url":"https://onlylabs.fyi/signals/7cb6383a-3088-4f3d-a2d2-efb7843da1d5","json_url":"https://onlylabs.fyi/signals/7cb6383a-3088-4f3d-a2d2-efb7843da1d5/signal.json","generated_at":"2026-06-11T04:20:02.195235+00:00","org":{"slug":"ai21","name":"AI21 Labs","category":"neolab","category_label":"Neolab","dossier_url":"https://onlylabs.fyi/labs/ai21","dossier_json_url":"https://onlylabs.fyi/labs/ai21/dossier.json"},"related_urls":{"signal":"https://onlylabs.fyi/signals/7cb6383a-3088-4f3d-a2d2-efb7843da1d5","signal_json":"https://onlylabs.fyi/signals/7cb6383a-3088-4f3d-a2d2-efb7843da1d5/signal.json","source":"https://github.com/AI21Labs/pmi-masking","lab_dossier":"https://onlylabs.fyi/labs/ai21","lab_dossier_json":"https://onlylabs.fyi/labs/ai21/dossier.json","analysis":"https://onlylabs.fyi/analysis/ai21","analysis_json":"https://onlylabs.fyi/analysis/ai21/analysis.json","analysis_evidence_json":"https://onlylabs.fyi/analysis/ai21/evidence.json","category":"https://onlylabs.fyi/neolabs","category_json":"https://onlylabs.fyi/neolabs.json","category_feed":"https://onlylabs.fyi/neolabs/feed.xml","category_signals_json":"https://onlylabs.fyi/signals.json?category=neolab","topic":null,"topic_signals_json":null,"topic_feed":null,"data_business":null},"answer_pack":{"answer":"AI21 Labs published AI21Labs/pmi-masking. This repository signal exposes tooling, eval, infrastructure, or model-adjacent work before it may appear in a launch post. High-signal details: repo AI21Labs/pmi-masking · AI21Labs/pmi-masking Description: This repository includes the masking vocabulary used in the ICLR 2021 spotlight PMI-Masking paper Stars: 14 Forks: 3 Open issues: 1.... onlylabs links this event to 1 captured evidence page and 6 related repo signals.","signal_desk":"repos","source_context":{"source_url":"https://github.com/AI21Labs/pmi-masking","source_host":"github.com","occurred_at":"2021-07-07T13:35:35+00:00","first_seen_at":"2026-06-05T20:58:44.654825+00:00","date_source":"source","context":null},"context_markers":[{"label":"Lab","value":"AI21 Labs","source":"signal"},{"label":"Signal desk","value":"repos","source":"signal"},{"label":"Source host","value":"github.com","source":"source"},{"label":"Repository","value":"AI21Labs/pmi-masking","source":"source"},{"label":"Stars","value":"14","source":"traction"},{"label":"Watch term","value":"Eval methodology","source":"evidence"},{"label":"Watch term","value":"Infrastructure","source":"evidence"},{"label":"Watch term","value":"Safety and alignment","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://github.com/AI21Labs/pmi-masking"],"related_signals":6,"has_source_url":true,"latest_page_fetched_at":"2026-06-11T04:20:02.195235+00:00"},"data_business":{"matches":false,"lanes":[],"matched_terms":[],"score":null,"reason":null},"agent_handoff":{"signal_json":"https://onlylabs.fyi/signals/7cb6383a-3088-4f3d-a2d2-efb7843da1d5/signal.json","dossier_json":"https://onlylabs.fyi/labs/ai21/dossier.json","analysis_json":"https://onlylabs.fyi/analysis/ai21/analysis.json","analysis_evidence_json":"https://onlylabs.fyi/analysis/ai21/evidence.json","topic_signals_json":null,"topic_feed":null,"category_signals_json":"https://onlylabs.fyi/signals.json?category=neolab","data_radar_json":null,"opportunities_json":null},"analysis_playbook":{"objective":"Turn new repository signals into early evidence of tooling, eval, infrastructure, model-adjacent, or product work before it appears in polished launch channels.","evidence_focus":["repo name","owner","description","language","stars","source URL","first seen time","data, eval, infra, safety, and product terms"],"extraction_questions":["What technical area does this repository expose?","Does the repo imply eval, data, infrastructure, agent, or deployment work?","Is the repo new evidence for a lab direction that is not yet in writing or releases?","Which related signals should an analyst inspect next?"],"signal_questions":["What does this new repository reveal before a formal announcement exists?","What technical area does this repository expose?","Does the repo imply eval, data, infrastructure, agent, or deployment work?","Do the 6 related repo signals show a repeated pattern?"],"output_fields":["org","repo","technical_theme","evidence_url"],"data_business_relevance":"Data-business lane extraction is scoped to frontier labs; for this category, interpret the repository as source-grounded category strategy evidence.","required_sources":[{"label":"signal_json","url":"https://onlylabs.fyi/signals/7cb6383a-3088-4f3d-a2d2-efb7843da1d5/signal.json","required":true},{"label":"source","url":"https://github.com/AI21Labs/pmi-masking","required":true},{"label":"dossier_json","url":"https://onlylabs.fyi/labs/ai21/dossier.json","required":true},{"label":"analysis_evidence_json","url":"https://onlylabs.fyi/analysis/ai21/evidence.json","required":true},{"label":"topic_signals_json","url":null,"required":false},{"label":"data_radar_json","url":null,"required":false}],"expected_output":["one-paragraph source-grounded interpretation","category-specific 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 AI21 Labs's repo signal \"AI21Labs/pmi-masking\" for neolab strategy."},"semantic_triples":[{"subject":"AI21 Labs","predicate":"published repo","object":"AI21Labs/pmi-masking","text":"AI21 Labs published repo AI21Labs/pmi-masking."},{"subject":"AI21Labs/pmi-masking","predicate":"is classified as","object":"repo signal","text":"AI21Labs/pmi-masking is classified as repo signal."},{"subject":"AI21Labs/pmi-masking","predicate":"belongs to","object":"repos desk","text":"AI21Labs/pmi-masking belongs to repos desk."},{"subject":"AI21Labs/pmi-masking","predicate":"has evidence coverage","object":"1 captured evidence page","text":"AI21Labs/pmi-masking has evidence coverage 1 captured evidence page."},{"subject":"AI21Labs/pmi-masking","predicate":"has captured page count","object":"1","text":"AI21Labs/pmi-masking has captured page count 1."},{"subject":"AI21Labs/pmi-masking","predicate":"has readable page count","object":"1","text":"AI21Labs/pmi-masking has readable page count 1."},{"subject":"AI21Labs/pmi-masking","predicate":"has related signal count","object":"6","text":"AI21Labs/pmi-masking has related signal count 6."},{"subject":"AI21Labs/pmi-masking","predicate":"has analysis playbook objective","object":"Turn new repository signals into early evidence of tooling, eval, infrastructure, model-adjacent, or product work before it appears in polished launch channels.","text":"AI21Labs/pmi-masking has analysis playbook objective Turn new repository signals into early evidence of tooling, eval, infrastructure, model-adjacent, or product work before it appears in polished launch channels.."},{"subject":"AI21Labs/pmi-masking","predicate":"has source host","object":"github.com","text":"AI21Labs/pmi-masking has source host github.com."},{"subject":"AI21Labs/pmi-masking","predicate":"has lab","object":"AI21 Labs","text":"AI21Labs/pmi-masking has lab AI21 Labs."},{"subject":"AI21Labs/pmi-masking","predicate":"has signal desk","object":"repos","text":"AI21Labs/pmi-masking has signal desk repos."},{"subject":"AI21Labs/pmi-masking","predicate":"has source host","object":"github.com","text":"AI21Labs/pmi-masking has source host github.com."},{"subject":"AI21Labs/pmi-masking","predicate":"has repository","object":"AI21Labs/pmi-masking","text":"AI21Labs/pmi-masking has repository AI21Labs/pmi-masking."},{"subject":"AI21Labs/pmi-masking","predicate":"has stars","object":"14","text":"AI21Labs/pmi-masking has stars 14."},{"subject":"AI21Labs/pmi-masking","predicate":"has watch term","object":"Eval methodology","text":"AI21Labs/pmi-masking has watch term Eval methodology."},{"subject":"AI21Labs/pmi-masking","predicate":"has watch term","object":"Infrastructure","text":"AI21Labs/pmi-masking has watch term Infrastructure."},{"subject":"AI21Labs/pmi-masking","predicate":"has watch term","object":"Safety and alignment","text":"AI21Labs/pmi-masking has watch term Safety and alignment."}]},"intelligence":{"signal_desk":"repos","answer":"AI21 Labs published AI21Labs/pmi-masking. This repository signal exposes tooling, eval, infrastructure, or model-adjacent work before it may appear in a launch post. High-signal details: repo AI21Labs/pmi-masking · AI21Labs/pmi-masking Description: This repository includes the masking vocabulary used in the ICLR 2021 spotlight PMI-Masking paper Stars: 14 Forks: 3 Open issues: 1.... onlylabs links this event to 1 captured evidence page and 6 related repo signals.","semantic_triples":[{"subject":"AI21 Labs","predicate":"published repo","object":"AI21Labs/pmi-masking","text":"AI21 Labs published repo AI21Labs/pmi-masking."},{"subject":"AI21Labs/pmi-masking","predicate":"is classified as","object":"repo signal","text":"AI21Labs/pmi-masking is classified as repo signal."},{"subject":"AI21Labs/pmi-masking","predicate":"belongs to","object":"repos desk","text":"AI21Labs/pmi-masking belongs to repos desk."},{"subject":"AI21Labs/pmi-masking","predicate":"has evidence coverage","object":"1 captured evidence page","text":"AI21Labs/pmi-masking has evidence coverage 1 captured evidence page."}]},"signal":{"id":"7cb6383a-3088-4f3d-a2d2-efb7843da1d5","url":"https://onlylabs.fyi/signals/7cb6383a-3088-4f3d-a2d2-efb7843da1d5","json_url":"https://onlylabs.fyi/signals/7cb6383a-3088-4f3d-a2d2-efb7843da1d5/signal.json","source_url":"https://github.com/AI21Labs/pmi-masking","title":"AI21Labs/pmi-masking","summary":"AI21 Labs published a new repository. onlylabs watches repos for tooling, eval, infra, and model-adjacent work.","context":null,"kind":{"key":"repo_new","label":"Repo"},"org":{"slug":"ai21","name":"AI21 Labs","category":"neolab"},"occurred_at":"2021-07-07T13:35:35+00:00","first_seen_at":"2026-06-05T20:58:44.654825+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://github.com/AI21Labs/pmi-masking"]},"facets":{"repo":"AI21Labs/pmi-masking"},"traction":{"github_stars":14,"hn_points":null,"hn_comments":null,"hn_story_id":null,"hf_downloads":null,"hf_likes":null},"data_radar":null},"primary_evidence_page":{"url":"https://github.com/AI21Labs/pmi-masking","final_url":"https://github.com/AI21Labs/pmi-masking","title":"AI21Labs/pmi-masking repository metadata","http_status":200,"content_type":"application/json","capture_method":"plain","fetched_at":"2026-06-11T04:20:02.195235+00:00","bytes":12032,"raw_path":"2b48c48d649178825237cf6ab9c019ea8380c8fb7edd02b05981e417454bd3c7.json","content_hash":"fe7b86d2460f329b70c1e80304082d81d93b033e8aa707a94ae14c9583b40388","excerpt_chars":1200,"truncated":true,"excerpt":"AI21Labs/pmi-masking Description: This repository includes the masking vocabulary used in the ICLR 2021 spotlight PMI-Masking paper Stars: 14 Forks: 3 Open issues: 1 Created: 2021-07-07T13:35:35Z Pushed: 2021-08-09T14:03:05Z Default branch: main Fork: no Archived: yes README: PMI-Masking This repository includes the list of masked spans (the masking vocabulary) that is used in the [ICLR 2021 spotlight](https://iclr.cc/virtual/2021/spotlight/3496) PMI-Masking [paper](https://openreview.net/forum?id=3Aoft6NWFej), overviewed in this [blogpost](https://www.ai21.com/blog/pmi-masking). Below, * Section 1 provides the list construction details. * Section 2 details the PMI-Masking method for bidirectional masked language models, which essentially treats all of the spans in the provided lists as units for masking. * Section 3 provides some scores of recently trained models on the [SQuAD2.0](https://rajpurkar.github.io/SQuAD-explorer/) and [RACE](https://www.cs.cmu.edu/~glai1/data/race/) benchmarks. Section 1: How we constructed the masking vocabulary Given a pretraining corpus, we use its n-gram co-occurrence statistics to compile our masking vocabulary. We consider word n-grams of lengths..."},"evidence_pages":[{"url":"https://github.com/AI21Labs/pmi-masking","final_url":"https://github.com/AI21Labs/pmi-masking","title":"AI21Labs/pmi-masking repository metadata","http_status":200,"content_type":"application/json","capture_method":"plain","fetched_at":"2026-06-11T04:20:02.195235+00:00","bytes":12032,"raw_path":"2b48c48d649178825237cf6ab9c019ea8380c8fb7edd02b05981e417454bd3c7.json","content_hash":"fe7b86d2460f329b70c1e80304082d81d93b033e8aa707a94ae14c9583b40388","excerpt_chars":1200,"truncated":true,"excerpt":"AI21Labs/pmi-masking Description: This repository includes the masking vocabulary used in the ICLR 2021 spotlight PMI-Masking paper Stars: 14 Forks: 3 Open issues: 1 Created: 2021-07-07T13:35:35Z Pushed: 2021-08-09T14:03:05Z Default branch: main Fork: no Archived: yes README: PMI-Masking This repository includes the list of masked spans (the masking vocabulary) that is used in the [ICLR 2021 spotlight](https://iclr.cc/virtual/2021/spotlight/3496) PMI-Masking [paper](https://openreview.net/forum?id=3Aoft6NWFej), overviewed in this [blogpost](https://www.ai21.com/blog/pmi-masking). Below, * Section 1 provides the list construction details. * Section 2 details the PMI-Masking method for bidirectional masked language models, which essentially treats all of the spans in the provided lists as units for masking. * Section 3 provides some scores of recently trained models on the [SQuAD2.0](https://rajpurkar.github.io/SQuAD-explorer/) and [RACE](https://www.cs.cmu.edu/~glai1/data/race/) benchmarks. Section 1: How we constructed the masking vocabulary Given a pretraining corpus, we use its n-gram co-occurrence statistics to compile our masking vocabulary. We consider word n-grams of lengths..."}],"related_signals":[{"id":"a8d5a050-6c2c-48fc-8704-64a597e69658","url":"https://onlylabs.fyi/signals/a8d5a050-6c2c-48fc-8704-64a597e69658","source_url":"https://github.com/AI21Labs/pre-commit-hadolint","title":"AI21Labs/pre-commit-hadolint","context":"Python","kind":{"key":"repo_new","label":"Repo"},"org":{"slug":"ai21","name":"AI21 Labs","category":"neolab"},"occurred_at":"2026-05-12T08:34:02+00:00","first_seen_at":"2026-06-05T20:58:44.654825+00:00","date_source":"source"},{"id":"7beb24f5-17a5-4573-92ef-792037a0120f","url":"https://onlylabs.fyi/signals/7beb24f5-17a5-4573-92ef-792037a0120f","source_url":"https://github.com/AI21Labs/multi-window-chunk-size","title":"AI21Labs/multi-window-chunk-size","context":"Jupyter Notebook","kind":{"key":"repo_new","label":"Repo"},"org":{"slug":"ai21","name":"AI21 Labs","category":"neolab"},"occurred_at":"2026-01-26T12:02:59+00:00","first_seen_at":"2026-06-05T20:58:44.654825+00:00","date_source":"source"},{"id":"ac9e3f80-ea99-4dba-8929-da82435a3a34","url":"https://onlylabs.fyi/signals/ac9e3f80-ea99-4dba-8929-da82435a3a34","source_url":"https://github.com/AI21Labs/ai21-typescript","title":"AI21Labs/ai21-typescript","context":"TypeScript","kind":{"key":"repo_new","label":"Repo"},"org":{"slug":"ai21","name":"AI21 Labs","category":"neolab"},"occurred_at":"2024-11-05T09:44:25+00:00","first_seen_at":"2026-06-05T20:58:44.654825+00:00","date_source":"source"},{"id":"653c8c77-1a4f-4e50-8504-f5f596fb5be3","url":"https://onlylabs.fyi/signals/653c8c77-1a4f-4e50-8504-f5f596fb5be3","source_url":"https://github.com/AI21Labs/salt","title":"AI21Labs/salt","context":"Python","kind":{"key":"repo_new","label":"Repo"},"org":{"slug":"ai21","name":"AI21 Labs","category":"neolab"},"occurred_at":"2024-07-21T09:13:44+00:00","first_seen_at":"2026-06-05T20:58:44.654825+00:00","date_source":"source"},{"id":"d4083267-defa-4b38-8f7f-90ee756c2801","url":"https://onlylabs.fyi/signals/d4083267-defa-4b38-8f7f-90ee756c2801","source_url":"https://github.com/AI21Labs/AI21-Industry-Samples","title":"AI21Labs/AI21-Industry-Samples","context":"Jupyter Notebook","kind":{"key":"repo_new","label":"Repo"},"org":{"slug":"ai21","name":"AI21 Labs","category":"neolab"},"occurred_at":"2024-02-02T14:30:39+00:00","first_seen_at":"2026-06-05T20:58:44.654825+00:00","date_source":"source"},{"id":"3ceb6253-41fc-44c7-a33e-62f570e3922b","url":"https://onlylabs.fyi/signals/3ceb6253-41fc-44c7-a33e-62f570e3922b","source_url":"https://github.com/AI21Labs/ai21-python","title":"AI21Labs/ai21-python","context":"Python","kind":{"key":"repo_new","label":"Repo"},"org":{"slug":"ai21","name":"AI21 Labs","category":"neolab"},"occurred_at":"2023-12-10T09:20:15+00:00","first_seen_at":"2026-06-05T20:58:44.654825+00:00","date_source":"source"}]}