{"schema_version":"onlylabs.public_signal.v1","title":"Tencent Hunyuan Repo: Tencent-Hunyuan/Simple-Attention-Sparsification","description":"Tencent Hunyuan repo signal with public source context, captured evidence pages, related signals, and data-business radar classification.","url":"https://onlylabs.fyi/signals/5bac51e6-42f2-421b-9727-c35ba4760d20","json_url":"https://onlylabs.fyi/signals/5bac51e6-42f2-421b-9727-c35ba4760d20/signal.json","generated_at":"2026-10-12T23:50:11.128Z","evidence_latest_fetched_at":"2026-09-14T07:03:10.581366+00:00","signal_first_seen_at":"2026-09-14T07:00:58.317983+00:00","org":{"slug":"tencent-hunyuan","name":"Tencent Hunyuan","category":"frontier-lab","category_label":"Frontier lab","dossier_url":"https://onlylabs.fyi/labs/tencent-hunyuan","dossier_json_url":"https://onlylabs.fyi/labs/tencent-hunyuan/dossier.json"},"related_urls":{"signal":"https://onlylabs.fyi/signals/5bac51e6-42f2-421b-9727-c35ba4760d20","signal_json":"https://onlylabs.fyi/signals/5bac51e6-42f2-421b-9727-c35ba4760d20/signal.json","source":"https://github.com/Tencent-Hunyuan/Simple-Attention-Sparsification","lab_dossier":"https://onlylabs.fyi/labs/tencent-hunyuan","lab_dossier_json":"https://onlylabs.fyi/labs/tencent-hunyuan/dossier.json","analysis":"https://onlylabs.fyi/analysis/tencent-hunyuan","analysis_json":"https://onlylabs.fyi/analysis/tencent-hunyuan/analysis.json","analysis_evidence_json":"https://onlylabs.fyi/analysis/tencent-hunyuan/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":null,"topic_signals_json":null,"topic_feed":null,"data_business":null},"answer_pack":{"answer":"Tencent Hunyuan published Tencent-Hunyuan/Simple-Attention-Sparsification (Python). This repository signal exposes tooling, eval, infrastructure, or model-adjacent work before it may appear in a launch post. High-signal details: repo Tencent-Hunyuan/Simple-Attention-Sparsification · language Python. onlylabs links this event to 1 captured evidence page and 6 related repo signals.","signal_desk":"repos","source_context":{"source_url":"https://github.com/Tencent-Hunyuan/Simple-Attention-Sparsification","source_host":"github.com","occurred_at":"2026-09-10T16:50:04+00:00","first_seen_at":"2026-09-14T07:00:58.317983+00:00","date_source":"source","context":"Python"},"context_markers":[{"label":"Lab","value":"Tencent Hunyuan","source":"signal"},{"label":"Signal desk","value":"repos","source":"signal"},{"label":"Source host","value":"github.com","source":"source"},{"label":"Repository","value":"Tencent-Hunyuan/Simple-Attention-Sparsification","source":"source"},{"label":"Language","value":"Python","source":"source"},{"label":"Stars","value":"66","source":"traction"},{"label":"Watch term","value":"RL environments","source":"evidence"},{"label":"Watch term","value":"Eval methodology","source":"evidence"},{"label":"Watch term","value":"Data pipeline","source":"evidence"},{"label":"Watch term","value":"Infrastructure","source":"evidence"},{"label":"Watch term","value":"Agents and tool use","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/Tencent-Hunyuan/Simple-Attention-Sparsification"],"related_signals":6,"has_source_url":true,"latest_page_fetched_at":"2026-09-14T07:03:10.581366+00:00"},"data_business":{"matches":false,"lanes":[],"matched_terms":[],"score":null,"reason":null},"agent_handoff":{"signal_json":"https://onlylabs.fyi/signals/5bac51e6-42f2-421b-9727-c35ba4760d20/signal.json","dossier_json":"https://onlylabs.fyi/labs/tencent-hunyuan/dossier.json","analysis_json":"https://onlylabs.fyi/analysis/tencent-hunyuan/analysis.json","analysis_evidence_json":"https://onlylabs.fyi/analysis/tencent-hunyuan/evidence.json","topic_signals_json":null,"topic_feed":null,"category_signals_json":"https://onlylabs.fyi/signals.json","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","data_business_lane","evidence_url"],"data_business_relevance":"New repositories can expose organization build priorities early, especially around internal tooling, eval infrastructure, data systems, deployment, and agent workflows.","required_sources":[{"label":"signal_json","url":"https://onlylabs.fyi/signals/5bac51e6-42f2-421b-9727-c35ba4760d20/signal.json","required":true},{"label":"source","url":"https://github.com/Tencent-Hunyuan/Simple-Attention-Sparsification","required":true},{"label":"dossier_json","url":"https://onlylabs.fyi/labs/tencent-hunyuan/dossier.json","required":true},{"label":"analysis_evidence_json","url":"https://onlylabs.fyi/analysis/tencent-hunyuan/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 Tencent Hunyuan's repo signal \"Tencent-Hunyuan/Simple-Attention-Sparsification\" for frontier lab strategy."},"semantic_triples":[{"subject":"Tencent Hunyuan","predicate":"published repo","object":"Tencent-Hunyuan/Simple-Attention-Sparsification","text":"Tencent Hunyuan published repo Tencent-Hunyuan/Simple-Attention-Sparsification."},{"subject":"Tencent-Hunyuan/Simple-Attention-Sparsification","predicate":"is classified as","object":"repo signal","text":"Tencent-Hunyuan/Simple-Attention-Sparsification is classified as repo signal."},{"subject":"Tencent-Hunyuan/Simple-Attention-Sparsification","predicate":"belongs to","object":"repos desk","text":"Tencent-Hunyuan/Simple-Attention-Sparsification belongs to repos desk."},{"subject":"Tencent-Hunyuan/Simple-Attention-Sparsification","predicate":"has context","object":"Python","text":"Tencent-Hunyuan/Simple-Attention-Sparsification has context Python."},{"subject":"Tencent-Hunyuan/Simple-Attention-Sparsification","predicate":"has evidence coverage","object":"1 captured evidence page","text":"Tencent-Hunyuan/Simple-Attention-Sparsification has evidence coverage 1 captured evidence page."},{"subject":"Tencent-Hunyuan/Simple-Attention-Sparsification","predicate":"has captured page count","object":"1","text":"Tencent-Hunyuan/Simple-Attention-Sparsification has captured page count 1."},{"subject":"Tencent-Hunyuan/Simple-Attention-Sparsification","predicate":"has readable page count","object":"1","text":"Tencent-Hunyuan/Simple-Attention-Sparsification has readable page count 1."},{"subject":"Tencent-Hunyuan/Simple-Attention-Sparsification","predicate":"has related signal count","object":"6","text":"Tencent-Hunyuan/Simple-Attention-Sparsification has related signal count 6."},{"subject":"Tencent-Hunyuan/Simple-Attention-Sparsification","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":"Tencent-Hunyuan/Simple-Attention-Sparsification 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":"Tencent-Hunyuan/Simple-Attention-Sparsification","predicate":"has source host","object":"github.com","text":"Tencent-Hunyuan/Simple-Attention-Sparsification has source host github.com."},{"subject":"Tencent-Hunyuan/Simple-Attention-Sparsification","predicate":"has lab","object":"Tencent Hunyuan","text":"Tencent-Hunyuan/Simple-Attention-Sparsification has lab Tencent Hunyuan."},{"subject":"Tencent-Hunyuan/Simple-Attention-Sparsification","predicate":"has signal desk","object":"repos","text":"Tencent-Hunyuan/Simple-Attention-Sparsification has signal desk repos."},{"subject":"Tencent-Hunyuan/Simple-Attention-Sparsification","predicate":"has source host","object":"github.com","text":"Tencent-Hunyuan/Simple-Attention-Sparsification has source host github.com."},{"subject":"Tencent-Hunyuan/Simple-Attention-Sparsification","predicate":"has repository","object":"Tencent-Hunyuan/Simple-Attention-Sparsification","text":"Tencent-Hunyuan/Simple-Attention-Sparsification has repository Tencent-Hunyuan/Simple-Attention-Sparsification."},{"subject":"Tencent-Hunyuan/Simple-Attention-Sparsification","predicate":"has language","object":"Python","text":"Tencent-Hunyuan/Simple-Attention-Sparsification has language Python."},{"subject":"Tencent-Hunyuan/Simple-Attention-Sparsification","predicate":"has stars","object":"66","text":"Tencent-Hunyuan/Simple-Attention-Sparsification has stars 66."},{"subject":"Tencent-Hunyuan/Simple-Attention-Sparsification","predicate":"has watch term","object":"RL environments","text":"Tencent-Hunyuan/Simple-Attention-Sparsification has watch term RL environments."},{"subject":"Tencent-Hunyuan/Simple-Attention-Sparsification","predicate":"has watch term","object":"Eval methodology","text":"Tencent-Hunyuan/Simple-Attention-Sparsification has watch term Eval methodology."}]},"intelligence":{"signal_desk":"repos","answer":"Tencent Hunyuan published Tencent-Hunyuan/Simple-Attention-Sparsification (Python). This repository signal exposes tooling, eval, infrastructure, or model-adjacent work before it may appear in a launch post. High-signal details: repo Tencent-Hunyuan/Simple-Attention-Sparsification · language Python. onlylabs links this event to 1 captured evidence page and 6 related repo signals.","semantic_triples":[{"subject":"Tencent Hunyuan","predicate":"published repo","object":"Tencent-Hunyuan/Simple-Attention-Sparsification","text":"Tencent Hunyuan published repo Tencent-Hunyuan/Simple-Attention-Sparsification."},{"subject":"Tencent-Hunyuan/Simple-Attention-Sparsification","predicate":"is classified as","object":"repo signal","text":"Tencent-Hunyuan/Simple-Attention-Sparsification is classified as repo signal."},{"subject":"Tencent-Hunyuan/Simple-Attention-Sparsification","predicate":"belongs to","object":"repos desk","text":"Tencent-Hunyuan/Simple-Attention-Sparsification belongs to repos desk."},{"subject":"Tencent-Hunyuan/Simple-Attention-Sparsification","predicate":"has context","object":"Python","text":"Tencent-Hunyuan/Simple-Attention-Sparsification has context Python."},{"subject":"Tencent-Hunyuan/Simple-Attention-Sparsification","predicate":"has evidence coverage","object":"1 captured evidence page","text":"Tencent-Hunyuan/Simple-Attention-Sparsification has evidence coverage 1 captured evidence page."}]},"signal":{"id":"5bac51e6-42f2-421b-9727-c35ba4760d20","url":"https://onlylabs.fyi/signals/5bac51e6-42f2-421b-9727-c35ba4760d20","json_url":"https://onlylabs.fyi/signals/5bac51e6-42f2-421b-9727-c35ba4760d20/signal.json","source_url":"https://github.com/Tencent-Hunyuan/Simple-Attention-Sparsification","title":"Tencent-Hunyuan/Simple-Attention-Sparsification","summary":"Tencent Hunyuan published a new repository. onlylabs watches repos for tooling, eval, infra, and model-adjacent work.","context":"Python","kind":{"key":"repo_new","label":"Repo"},"org":{"slug":"tencent-hunyuan","name":"Tencent Hunyuan","category":"frontier-lab"},"occurred_at":"2026-09-10T16:50:04+00:00","first_seen_at":"2026-09-14T07:00:58.317983+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/Tencent-Hunyuan/Simple-Attention-Sparsification"]},"facets":{"repo":"Tencent-Hunyuan/Simple-Attention-Sparsification","language":"Python"},"traction":{"github_stars":66,"hn_points":null,"hn_comments":null,"hn_story_id":null,"hf_downloads":null,"hf_likes":null},"data_radar":null},"primary_evidence_page":{"is_primary":true,"source_match":true,"url":"https://github.com/Tencent-Hunyuan/Simple-Attention-Sparsification","final_url":"https://github.com/Tencent-Hunyuan/Simple-Attention-Sparsification","title":"Tencent-Hunyuan/Simple-Attention-Sparsification repository metadata","http_status":200,"content_type":"application/json","capture_method":"plain","fetched_at":"2026-09-14T07:03:10.581366+00:00","bytes":15718,"raw_path":"bdbe45871bf992a80a0d1adc1d112c750310da3c294251647dcdd3178174ec10.json","content_hash":"a7920dd237ddd8db2c50be72ed23954e9c5181d0d691c55a5e26859386e37548","excerpt_chars":1200,"truncated":true,"excerpt":"Tencent-Hunyuan/Simple-Attention-Sparsification Description: Code for our resarch paper \"SAS: Simple Attention Sparsification via End-to-End Optimization of Context Ranking\" Language: Python Stars: 26 Forks: 0 Open issues: 0 Created: 2026-09-10T16:50:04Z Pushed: 2026-09-14T06:45:59Z Default branch: release Fork: no Archived: no README: SAS: Simple Attention Sparsification via End-to-End Optimization of Context Ranking This repository provides the official implementation of [**SAS**](https://arxiv.org/abs/2609.13141). **SAS** is a gated sparse-attention mechanism that learns, per query, which blocks to attend to — optimizing the context ranking **end-to-end** with the language modeling loss instead of distilling the original model's dense attention. ![Overview of Simple Attention Sparsification](assets/method.png) ***(a) Gradient blockage in discrete selection.** Standard sparse attention relies on discrete Top-$K$ selection, which prevents gradients from the language-modeling loss from reaching the selector. Consequently, existing methods typically train the selector through auxiliary distillation objectives or hand-designed heuristics rather than optimizing context ranking..."},"evidence_pages":[],"related_signals":[{"id":"294f6a47-4152-41eb-95f4-78b4749043bd","url":"https://onlylabs.fyi/signals/294f6a47-4152-41eb-95f4-78b4749043bd","source_url":"https://github.com/Tencent-Hunyuan/ExplorationBench","title":"Tencent-Hunyuan/ExplorationBench","context":"Python","kind":{"key":"repo_new","label":"Repo"},"org":{"slug":"tencent-hunyuan","name":"Tencent Hunyuan","category":"frontier-lab"},"occurred_at":"2026-09-30T10:51:12+00:00","first_seen_at":"2026-10-08T07:00:53.450555+00:00","date_source":"source"},{"id":"0f5d69f7-cdc1-43dd-bd8e-bc82b53743b4","url":"https://onlylabs.fyi/signals/0f5d69f7-cdc1-43dd-bd8e-bc82b53743b4","source_url":"https://github.com/Tencent-Hunyuan/Prism","title":"Tencent-Hunyuan/Prism","context":"Python","kind":{"key":"repo_new","label":"Repo"},"org":{"slug":"tencent-hunyuan","name":"Tencent Hunyuan","category":"frontier-lab"},"occurred_at":"2026-09-21T11:31:58+00:00","first_seen_at":"2026-10-06T07:00:51.136324+00:00","date_source":"source"},{"id":"0d6b8fde-f4c8-4b43-9f13-38c0cbdae5c1","url":"https://onlylabs.fyi/signals/0d6b8fde-f4c8-4b43-9f13-38c0cbdae5c1","source_url":"https://github.com/Tencent-Hunyuan/evolve-scaler","title":"Tencent-Hunyuan/evolve-scaler","context":"HTML","kind":{"key":"repo_new","label":"Repo"},"org":{"slug":"tencent-hunyuan","name":"Tencent Hunyuan","category":"frontier-lab"},"occurred_at":"2026-09-08T07:17:34+00:00","first_seen_at":"2026-09-09T07:00:50.393695+00:00","date_source":"source"},{"id":"61479c69-db13-4d0f-9ffc-d1053b225295","url":"https://onlylabs.fyi/signals/61479c69-db13-4d0f-9ffc-d1053b225295","source_url":"https://github.com/Tencent-Hunyuan/Hy4-preview","title":"Tencent-Hunyuan/Hy4-preview","context":"Python","kind":{"key":"repo_new","label":"Repo"},"org":{"slug":"tencent-hunyuan","name":"Tencent Hunyuan","category":"frontier-lab"},"occurred_at":"2026-08-27T09:27:34+00:00","first_seen_at":"2026-08-28T07:00:48.344557+00:00","date_source":"source"},{"id":"b900c897-bcb5-4662-b077-29f54a6d31dd","url":"https://onlylabs.fyi/signals/b900c897-bcb5-4662-b077-29f54a6d31dd","source_url":"https://github.com/Tencent-Hunyuan/AuK","title":"Tencent-Hunyuan/AuK","context":"Python","kind":{"key":"repo_new","label":"Repo"},"org":{"slug":"tencent-hunyuan","name":"Tencent Hunyuan","category":"frontier-lab"},"occurred_at":"2026-08-19T01:39:24+00:00","first_seen_at":"2026-09-09T07:00:50.393695+00:00","date_source":"source"},{"id":"8edff6ed-fb3f-4b4e-851d-1d9a07baeabe","url":"https://onlylabs.fyi/signals/8edff6ed-fb3f-4b4e-851d-1d9a07baeabe","source_url":"https://github.com/Tencent-Hunyuan/RepoProbe","title":"Tencent-Hunyuan/RepoProbe","context":"Python","kind":{"key":"repo_new","label":"Repo"},"org":{"slug":"tencent-hunyuan","name":"Tencent Hunyuan","category":"frontier-lab"},"occurred_at":"2026-08-05T12:42:23+00:00","first_seen_at":"2026-08-06T07:00:53.111549+00:00","date_source":"source"}]}