{"schema_version":"onlylabs.public_signal.v1","title":"Tencent Hunyuan Model: tencent/Simple-Attention-Sparsification","description":"Tencent Hunyuan model signal with public source context, captured evidence pages, related signals, and data-business radar classification.","url":"https://onlylabs.fyi/signals/6440c29b-cdb1-4601-b785-887db64e2080","json_url":"https://onlylabs.fyi/signals/6440c29b-cdb1-4601-b785-887db64e2080/signal.json","generated_at":"2026-10-12T19:55:57.901Z","evidence_latest_fetched_at":"2026-09-15T07:20:26.518775+00:00","signal_first_seen_at":"2026-09-15T07:20:09.590751+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/6440c29b-cdb1-4601-b785-887db64e2080","signal_json":"https://onlylabs.fyi/signals/6440c29b-cdb1-4601-b785-887db64e2080/signal.json","source":"https://huggingface.co/tencent/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":"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":null},"answer_pack":{"answer":"Tencent Hunyuan published tencent/Simple-Attention-Sparsification. This model signal is evidence of what shipped on model infrastructure and how the release is positioned. High-signal details: 0 HF downloads · Technique to make transformer attention sparse and faster.. onlylabs links this event to 1 captured evidence page and 6 related model signals.","signal_desk":"releases","source_context":{"source_url":"https://huggingface.co/tencent/Simple-Attention-Sparsification","source_host":"huggingface.co","occurred_at":"2026-09-14T07:43:25+00:00","first_seen_at":"2026-09-15T07:20:09.590751+00:00","date_source":"source","context":null},"context_markers":[{"label":"Lab","value":"Tencent Hunyuan","source":"signal"},{"label":"Signal desk","value":"releases","source":"signal"},{"label":"Source host","value":"huggingface.co","source":"source"},{"label":"Task","value":"text-generation","source":"model"},{"label":"Downloads","value":"0","source":"traction"},{"label":"Likes","value":"13","source":"traction"},{"label":"Context","value":"Technique to make transformer attention sparse and faster.","source":"signal"},{"label":"Notability","value":"Tencent Hunyuan attention sparsification repo, likely substantive","source":"signal"},{"label":"Watch term","value":"Eval methodology","source":"evidence"},{"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/tencent/Simple-Attention-Sparsification/raw/main/README.md"],"related_signals":6,"has_source_url":true,"latest_page_fetched_at":"2026-09-15T07:20:26.518775+00:00"},"data_business":{"matches":false,"lanes":[],"matched_terms":[],"score":null,"reason":null},"agent_handoff":{"signal_json":"https://onlylabs.fyi/signals/6440c29b-cdb1-4601-b785-887db64e2080/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":"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":null,"opportunities_json":null},"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?","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/6440c29b-cdb1-4601-b785-887db64e2080/signal.json","required":true},{"label":"source","url":"https://huggingface.co/tencent/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":"https://onlylabs.fyi/topics/releases/signals.json","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 model signal \"tencent/Simple-Attention-Sparsification\" for frontier lab strategy."},"semantic_triples":[{"subject":"Tencent Hunyuan","predicate":"published model","object":"tencent/Simple-Attention-Sparsification","text":"Tencent Hunyuan published model tencent/Simple-Attention-Sparsification."},{"subject":"tencent/Simple-Attention-Sparsification","predicate":"is classified as","object":"model signal","text":"tencent/Simple-Attention-Sparsification is classified as model signal."},{"subject":"tencent/Simple-Attention-Sparsification","predicate":"belongs to","object":"releases desk","text":"tencent/Simple-Attention-Sparsification belongs to releases desk."},{"subject":"tencent/Simple-Attention-Sparsification","predicate":"has evidence coverage","object":"1 captured evidence page","text":"tencent/Simple-Attention-Sparsification has evidence coverage 1 captured evidence page."},{"subject":"tencent/Simple-Attention-Sparsification","predicate":"has captured page count","object":"1","text":"tencent/Simple-Attention-Sparsification has captured page count 1."},{"subject":"tencent/Simple-Attention-Sparsification","predicate":"has readable page count","object":"1","text":"tencent/Simple-Attention-Sparsification has readable page count 1."},{"subject":"tencent/Simple-Attention-Sparsification","predicate":"has related signal count","object":"6","text":"tencent/Simple-Attention-Sparsification has related signal count 6."},{"subject":"tencent/Simple-Attention-Sparsification","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":"tencent/Simple-Attention-Sparsification has analysis playbook objective Turn model cards, repository releases, and launch artifacts into a current view of what shipped and what changed.."},{"subject":"tencent/Simple-Attention-Sparsification","predicate":"has source host","object":"huggingface.co","text":"tencent/Simple-Attention-Sparsification has source host huggingface.co."},{"subject":"tencent/Simple-Attention-Sparsification","predicate":"has lab","object":"Tencent Hunyuan","text":"tencent/Simple-Attention-Sparsification has lab Tencent Hunyuan."},{"subject":"tencent/Simple-Attention-Sparsification","predicate":"has signal desk","object":"releases","text":"tencent/Simple-Attention-Sparsification has signal desk releases."},{"subject":"tencent/Simple-Attention-Sparsification","predicate":"has source host","object":"huggingface.co","text":"tencent/Simple-Attention-Sparsification has source host huggingface.co."},{"subject":"tencent/Simple-Attention-Sparsification","predicate":"has task","object":"text-generation","text":"tencent/Simple-Attention-Sparsification has task text-generation."},{"subject":"tencent/Simple-Attention-Sparsification","predicate":"has downloads","object":"0","text":"tencent/Simple-Attention-Sparsification has downloads 0."},{"subject":"tencent/Simple-Attention-Sparsification","predicate":"has likes","object":"13","text":"tencent/Simple-Attention-Sparsification has likes 13."},{"subject":"tencent/Simple-Attention-Sparsification","predicate":"has context","object":"Technique to make transformer attention sparse and faster.","text":"tencent/Simple-Attention-Sparsification has context Technique to make transformer attention sparse and faster.."},{"subject":"tencent/Simple-Attention-Sparsification","predicate":"has notability","object":"Tencent Hunyuan attention sparsification repo, likely substantive","text":"tencent/Simple-Attention-Sparsification has notability Tencent Hunyuan attention sparsification repo, likely substantive."}]},"intelligence":{"signal_desk":"releases","answer":"Tencent Hunyuan published tencent/Simple-Attention-Sparsification. This model signal is evidence of what shipped on model infrastructure and how the release is positioned. High-signal details: 0 HF downloads · Technique to make transformer attention sparse and faster.. onlylabs links this event to 1 captured evidence page and 6 related model signals.","semantic_triples":[{"subject":"Tencent Hunyuan","predicate":"published model","object":"tencent/Simple-Attention-Sparsification","text":"Tencent Hunyuan published model tencent/Simple-Attention-Sparsification."},{"subject":"tencent/Simple-Attention-Sparsification","predicate":"is classified as","object":"model signal","text":"tencent/Simple-Attention-Sparsification is classified as model signal."},{"subject":"tencent/Simple-Attention-Sparsification","predicate":"belongs to","object":"releases desk","text":"tencent/Simple-Attention-Sparsification belongs to releases desk."},{"subject":"tencent/Simple-Attention-Sparsification","predicate":"has evidence coverage","object":"1 captured evidence page","text":"tencent/Simple-Attention-Sparsification has evidence coverage 1 captured evidence page."}]},"signal":{"id":"6440c29b-cdb1-4601-b785-887db64e2080","url":"https://onlylabs.fyi/signals/6440c29b-cdb1-4601-b785-887db64e2080","json_url":"https://onlylabs.fyi/signals/6440c29b-cdb1-4601-b785-887db64e2080/signal.json","source_url":"https://huggingface.co/tencent/Simple-Attention-Sparsification","title":"tencent/Simple-Attention-Sparsification","summary":"Tencent Hunyuan published a model signal. onlylabs watches model cards, launch context, and traction.","context":null,"kind":{"key":"model_released","label":"Model"},"org":{"slug":"tencent-hunyuan","name":"Tencent Hunyuan","category":"frontier-lab"},"occurred_at":"2026-09-14T07:43:25+00:00","first_seen_at":"2026-09-15T07:20:09.590751+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/tencent/Simple-Attention-Sparsification/raw/main/README.md"]},"facets":{"pipeline_tag":"text-generation"},"traction":{"github_stars":null,"hn_points":null,"hn_comments":null,"hn_story_id":null,"hf_downloads":0,"hf_likes":13},"data_radar":null},"primary_evidence_page":{"is_primary":true,"source_match":true,"url":"https://huggingface.co/tencent/Simple-Attention-Sparsification/raw/main/README.md","final_url":"https://huggingface.co/tencent/Simple-Attention-Sparsification/raw/main/README.md","title":"tencent/Simple-Attention-Sparsification model card","http_status":200,"content_type":"text/plain; charset=utf-8","capture_method":"plain","fetched_at":"2026-09-15T07:20:26.518775+00:00","bytes":6639,"raw_path":"682b193714579048ee8681e9ae3f4573d8fcf8cf234b9e622fbb41e105962b95.md","content_hash":"383b9e9cbe1c57057b6e57dcd8ed159c6e5eb13a0d60db9851b474c55c4fd163","excerpt_chars":1200,"truncated":true,"excerpt":"Simple Attention Sparsification (SAS) **SAS** learns to rank and select KV blocks for each query. Unlike methods that train a sparse-attention selector by distilling dense attention scores, SAS adds continuous gates to the selected blocks so that the language-modeling loss can optimize context ranking end to end. Overview of SAS. Continuous gates provide a differentiable path through discrete Top-K block selection, enabling end-to-end context-ranking optimization with the language modeling loss. Released checkpoints | Directory | Base model | Gate parameters | Gate checkpoint size | |---|---|---:|---:| | `Qwen3-4B-AttnGates` | [Qwen3-4B](https://huggingface.co/Qwen/Qwen3-4B) | 33.0M | 64 MiB | | `Qwen3-8B-AttnGates` | [Qwen3-8B](https://huggingface.co/Qwen/Qwen3-8B) | 33.0M | 64 MiB | | `Qwen3-14B-AttnGates` | [Qwen3-14B](https://huggingface.co/Qwen/Qwen3-14B) | 42.0M | 81 MiB | Each directory is an SGLang-compatible `AttnGates` package containing: - `attn_gate_weights.pth`: the learned gate-router weights; - `config.json`: the base-model reference and `seerattn_*` configuration; - tokenizer and chat-template files copied from the corresponding Qwen3 model. > [!IMPORTANT] > These..."},"evidence_pages":[],"related_signals":[{"id":"d0b510c5-71a7-4d0c-b2ea-44f98d76ab48","url":"https://onlylabs.fyi/signals/d0b510c5-71a7-4d0c-b2ea-44f98d76ab48","source_url":"https://huggingface.co/tencent/Youtu-Parsing-Omni","title":"tencent/Youtu-Parsing-Omni","context":null,"kind":{"key":"model_released","label":"Model"},"org":{"slug":"tencent-hunyuan","name":"Tencent Hunyuan","category":"frontier-lab"},"occurred_at":"2026-10-09T06:42:18+00:00","first_seen_at":"2026-10-10T07:20:09.04386+00:00","date_source":"source"},{"id":"28e9ca58-48d6-4cc3-a16f-a5ee56ebfe62","url":"https://onlylabs.fyi/signals/28e9ca58-48d6-4cc3-a16f-a5ee56ebfe62","source_url":"https://huggingface.co/tencent/WeVisDoc-4B","title":"tencent/WeVisDoc-4B","context":null,"kind":{"key":"model_released","label":"Model"},"org":{"slug":"tencent-hunyuan","name":"Tencent Hunyuan","category":"frontier-lab"},"occurred_at":"2026-09-16T03:07:01+00:00","first_seen_at":"2026-09-18T07:20:09.161463+00:00","date_source":"source"},{"id":"6ec1f81f-3829-4fd0-a60f-4a5231834509","url":"https://onlylabs.fyi/signals/6ec1f81f-3829-4fd0-a60f-4a5231834509","source_url":"https://huggingface.co/tencent/WeVisDoc-2B","title":"tencent/WeVisDoc-2B","context":null,"kind":{"key":"model_released","label":"Model"},"org":{"slug":"tencent-hunyuan","name":"Tencent Hunyuan","category":"frontier-lab"},"occurred_at":"2026-09-16T03:06:43+00:00","first_seen_at":"2026-09-18T07:20:09.161463+00:00","date_source":"source"},{"id":"212aca2d-f9f3-4b1d-b70c-48d4c3277b19","url":"https://onlylabs.fyi/signals/212aca2d-f9f3-4b1d-b70c-48d4c3277b19","source_url":"https://huggingface.co/tencent/EVIE-8B","title":"tencent/EVIE-8B","context":null,"kind":{"key":"model_released","label":"Model"},"org":{"slug":"tencent-hunyuan","name":"Tencent Hunyuan","category":"frontier-lab"},"occurred_at":"2026-09-04T08:55:13+00:00","first_seen_at":"2026-09-08T07:20:09.149803+00:00","date_source":"source"},{"id":"27427871-48ae-4aa3-8899-963df7999016","url":"https://onlylabs.fyi/signals/27427871-48ae-4aa3-8899-963df7999016","source_url":"https://huggingface.co/tencent/EVIE-4.5B","title":"tencent/EVIE-4.5B","context":null,"kind":{"key":"model_released","label":"Model"},"org":{"slug":"tencent-hunyuan","name":"Tencent Hunyuan","category":"frontier-lab"},"occurred_at":"2026-09-04T08:54:45+00:00","first_seen_at":"2026-09-08T07:20:09.149803+00:00","date_source":"source"},{"id":"e36f778a-8545-4d78-b3fd-a9da4e9b5298","url":"https://onlylabs.fyi/signals/e36f778a-8545-4d78-b3fd-a9da4e9b5298","source_url":"https://huggingface.co/tencent/Hy4-preview","title":"tencent/Hy4-preview","context":null,"kind":{"key":"model_released","label":"Model"},"org":{"slug":"tencent-hunyuan","name":"Tencent Hunyuan","category":"frontier-lab"},"occurred_at":"2026-08-27T08:52:40+00:00","first_seen_at":"2026-08-28T07:20:09.545171+00:00","date_source":"source"}]}