{"slug":"technical-report","label":"Technical Report","item_count":3,"day_count":3,"source_count":2,"first_seen":"2026-10-30T02:07:22+00:00","last_updated":"2026-11-18T15:46:14+00:00","generated_at":"2026-11-20T00:13:34.686460+00:00","sources":["arxiv_cs_cl","search_cn_open_weight_labs"],"days":[{"date":"2026-10-30","items":[{"title":"Tencent Hunyuan Publishes Hy-Nomas Technical Report and Releases Hy-Nomas-Bench for World Models","url":"https://news.google.com/rss/articles/CBMipwFBVV95cUxOVlZ5am9WcWMtVE9JLUIzRkVUaDI5QS1vQ29VYWpwWVJQdjUtUWJPZGxNdmR1dDJTZ0xoc1ZIV04zQlY2M0NxdWxYMHhRTFJ3UUl5LXNCMVJ2ZTFNZW9COUZLT05OU1NQT1Y5cThXSVY3eXI2aUdpbUpqOG1UYUJITjFMdFhYeXczQ2l1SmJjWWNwZFExQnJTMm42Sm8zM2tXaHBON1FJVQ?oc=5","source":"search_cn_open_weight_labs","type":"release","summary_1line":"Tencent Hunyuan Publishes Hy-Nomas Technical Report and Releases Hy-Nomas-Bench for World Models Pandaily","sid":"80bf40e7e133130c","published":"2026-10-30T02:07:22+00:00","editor_note":"Opens the thread: a world-model report plus a new benchmark, reported secondhand by Pandaily."}]},{"date":"2026-11-09","items":[{"title":"MNN-Pocket-0.3B Technical Report","url":"http://arxiv.org/abs/2611.12311v1","source":"arxiv_cs_cl","type":"paper","summary_1line":"This technical report presents MNN-Pocket-0.3B, a compact 300M-parameter dense language model designed for efficient on-device deployment and supporting context lengths of up to 32K tokens. With an INT4 deployment foo...","why_it_matters":"Matches feed focus: evaluation.","sid":"2f7be7ac68b1edc1","published":"2026-11-09T12:33:26+00:00","editor_note":"Adds the small end of the range: a 300M on-device model with an INT4 path."}]},{"date":"2026-11-18","items":[{"title":"Clover-35B-A3B-Preview Technical Report","url":"http://arxiv.org/abs/2611.24636v1","source":"arxiv_cs_cl","type":"paper","summary_1line":"We introduce Clover-35B-A3B-Preview (Clover), a mixture-of-experts diffusion language model with 35 billion total parameters and 3 billion activated parameters, supporting thinking-mode generation. Clover adapts Gated...","why_it_matters":"Matches feed focus: agent, evaluation.","sid":"b0b6c263afe6d2d0","published":"2026-11-18T15:46:14+00:00","editor_note":"Latest entry: a sparse diffusion LM, a different decoding approach from the autoregressive norm."}]}],"editorial":{"tldr":"This is a loose thread of unrelated model technical reports, not one developing event. It opened with Tencent Hunyuan's Hy-Nomas world-model report and benchmark on Oct 30, then MNN-Pocket-0.3B, a 300M-parameter on-device model with 32K context, on Nov 9.","stale":false,"whats_new":"Clover-35B-A3B-Preview's report (Nov 18) describes a mixture-of-experts diffusion language model with 35B total and 3B active parameters that supports a thinking mode.","why_it_matters":"Each report shows a different serving shape: world models, sub-1B INT4 on-device models, and sparse diffusion LMs, each with its own inference cost and runtime needs.","take_for_builders":"Treat these as research reads. Before adopting one, check its weight license and runtime support, especially for Clover's diffusion decoding.","beats":[{"kicker":"WORLD MODELS","tone":"launch","headline":"Tencent Hunyuan publishes the Hy-Nomas report and Hy-Nomas-Bench","summary":"A lab report paired with a new benchmark for world models; coverage came via secondary press.","sids":["80bf40e7e133130c"]},{"kicker":"ON-DEVICE","tone":"rising","headline":"MNN-Pocket-0.3B targets on-device deployment at 300M parameters","summary":"A compact dense model with 32K context and an INT4 deployment path.","sids":["2f7be7ac68b1edc1"]},{"kicker":"NOW","tone":"now","headline":"Clover-35B-A3B-Preview: a sparse MoE diffusion LM with thinking mode","summary":"35B total and 3B active parameters, built on an adapted gated architecture per the abstract.","sids":["b0b6c263afe6d2d0"]}],"open_questions":["Are the Hy-Nomas, MNN-Pocket-0.3B, or Clover weights released under a license that permits commercial use?","Does Hy-Nomas-Bench gain adoption outside Tencent's own evaluations?","Do mainstream inference servers support Clover's diffusion decoding, or does it need a custom runtime?","Do independent evals reproduce MNN-Pocket-0.3B's reported quality at INT4?"],"generated_at":"2026-11-20T00:12:41Z"}}