{"slug":"deepseek-open-sources","label":"DeepSeek Open-sources","item_count":45,"day_count":17,"source_count":10,"first_seen":"2026-10-17T12:42:12+00:00","last_updated":"2026-11-06T15:21:03+00:00","via_scout":true,"generated_at":"2026-11-07T00:12:50.771141+00:00","sources":["arxiv_agent_systems_research","arxiv_cs_ai","arxiv_cs_cl","arxiv_cs_lg","arxiv_llm_reliability","hackernews_ai","latent_space","search_agent_engineering_news","search_cn_open_weight_labs","search_llm_ops_news"],"days":[{"date":"2026-10-17","items":[{"title":"DeepSeek Ties Its New V4.1-Flash-1017 Model Tighter to Its Own Harness","url":"https://news.google.com/rss/articles/CBMingFBVV95cUxQZUFkNm9JaDNLYzVRVTcwaDNTNldBMS1XbVZlQ3RpcnBrU3NNczNicTFRYmY2RkY2bmVyZENQZkVhYmRLUnpKMFZhTmQwamVNUlp2RzRmWFlNZU1jdUxnUlZ1WHBBVWhlVnR4ck5YRFhmaTZ4WWV0Q3JGUmFJLUJkWUI4TzdrWTFXMGJoTW9UR0oxNWs0N1VXRnNVRE9FUQ?oc=5","source":"search_cn_open_weight_labs","type":"news","summary_1line":"DeepSeek Ties Its New V4.1-Flash-1017 Model Tighter to Its Own Harness Startup Fortune","why_it_matters":"Matches feed focus: harness.","sid":"ce1f00045df6fba8","published":"2026-10-17T12:42:12+00:00","editor_note":"Model tied to DeepSeek's own harness"}]},{"date":"2026-10-18","items":[{"title":"New Co-Design Framework Speeds Up LLM Inference on Mixed Hardware by Taming Memory Bottlenecks","url":"https://news.google.com/rss/articles/CBMiuwFBVV95cUxOdm1kSExLR1hxSXpsdTVpSmtTTnVlQkdNRjhRZHUwR1pxU3ZyYkxxSUZINk95N1d5Z1lxSmpBS3ZWbHVmRWp1dmY3REx5bnNYSGRTblczTUw5VFdIaER0UWM4Q2hZRTRQVDdOazZ5YmJVT3puV3p5YlZWcTB5M2hOTHIwYWxzcGM2a2gwY0c2VmZSRTZzSnY1SUIwb0V3T01XbDBwaFZYUWpNbXFhRFlnc0QxSHhIMGIteDJJ?oc=5","source":"search_llm_ops_news","type":"news","summary_1line":"New Co-Design Framework Speeds Up LLM Inference on Mixed Hardware by Taming Memory Bottlenecks Bioengineer.org","sid":"db7388ca69770009","published":"2026-10-18T06:06:18+00:00"}]},{"date":"2026-10-20","items":[{"title":"llm-d 1.0 Advances Open, Cost-Efficient AI Inference for Production Scale","url":"https://news.google.com/rss/articles/CBMizgFBVV95cUxQRENMaFZVanUzZXQwV21aVnRYWVZnU3g5WFNNUVAzamZCMF9UamVzbDlUMGRLaXctNVIzNEptalYwX2NqWUxqbzA1SEJUMkZEQ1Z6NlkxaUJqVlVodExja0RxR1Q1Sk91R1NRUjUxUFhpZlR6bGpzUWNURVZMV25CTXJDTGhtVk9PSmFTdmpSc0FJeVZ4X2c5R0dRMWYxWjUtWGdCbDBTQlJQU1VTMlFUWnBkTXlpMUVIbDJlVndUT3VvSE5UWHJQdXc4cS01QQ?oc=5","source":"search_llm_ops_news","type":"news","summary_1line":"llm-d 1.0 Advances Open, Cost-Efficient AI Inference for Production Scale TradingView","sid":"0714a9511cb60f66","published":"2026-10-20T12:00:00+00:00"},{"title":"DeepSeek Unveils Open-Source Tool to Help Cut AI Operating Costs","url":"https://news.google.com/rss/articles/CBMitgFBVV95cUxNaTNwMlNFRlZ6SkZCMWtZaEJfUWdXbGxlQVBnSEZOZ0dVM1dLc240NTVQdnZTbjJqd2pRWVNpUUZVYldWZDlzdF8waFlUVlRRaW1uYllNRjdVNFZZSW9aeFhGQmRzS1p1VmhFUG9HbTMzaE5hTm1mOGFwSHp5VzlEbTEtc2ZsQTJ3amdCVjhkZUU5bkY3UDM2ajNRcGY4aDJBWjhFbGljNjlZaXI1X3gyX1gyYTV4QQ?oc=5","source":"search_cn_open_weight_labs","type":"news","summary_1line":"DeepSeek Unveils Open-Source Tool to Help Cut AI Operating Costs Bloomberg.com","sid":"6ca9e1ceeea5a770","published":"2026-10-20T12:45:23+00:00"},{"title":"DeepSeek is open-sourcing the software it uses to run its AI models at scale","url":"https://news.google.com/rss/articles/CBMif0FVX3lxTE1IRHJQZVBMMTdReU1BbHdBX3JfNnZ4MUxfUEhZQnhzcW9zbklPa0xZZmxHWlMxeUg5QXExYld5NS02ZU1iWUlMeFdadjVpTFk2eWxLS00yR3lUMm9obl9jbE43RF8wQjZjWVlQSWpuN2J0eWN2dFdxckJDWU9ISDQ?oc=5","source":"search_cn_open_weight_labs","type":"news","summary_1line":"DeepSeek is open-sourcing the software it uses to run its AI models at scale Quartz","sid":"be61663792fa0af3","published":"2026-10-20T13:36:46+00:00"},{"title":"DeepSeek Dials Up the Pressure on OpenAI and Anthropic With Open-Source Inference Tool","url":"https://news.google.com/rss/articles/CBMiuAFBVV95cUxOb01YenVDbE5LRnNkaWVVOUczR3hpMFlVYWlTRk9BX3NlMUp0QVBHdWlzVTBEcGlPMmZXdmRCaWtIaUthbmlfQml2bjFSY1NkNVRlSUx0LVBWeTh4WkhzYklCTnJmc0JraFZFQ3NDRUV4ME1jbUlEOVYtY3kwNjFacXNkVmgwZGlfeHZsVF8tTzJQVXlvSVdubnRCQXlLeEVKdlNVYWlOOV9BUzJEeHh0aGd0MUlkU01Z?oc=5","source":"search_cn_open_weight_labs","type":"news","summary_1line":"DeepSeek Dials Up the Pressure on OpenAI and Anthropic With Open-Source Inference Tool Gizmodo","sid":"861b6347e652747b","published":"2026-10-20T16:45:40+00:00"},{"title":"DeepSeek Open Sources Inference Tech to Lower AI Costs","url":"https://news.google.com/rss/articles/CBMisgFBVV95cUxQU1g5bE5jajR2Tjc3Rmx3cjVXeThXSHdGMkJXYUM4YWR1eEVlS2pwUUg2RFY0OHIwLWlnQU5zT19IcTFaMVJlVGQxTDRYUkFmTHJWQWZPZjNkZHRYRGd5Ym5QbjZvX3RLc3dWb0l3bmZVNEVjdzVNVXVsaWtES0owM1B4YjFoM1VoNE9oYk9Vbko1Mm03YVhMaWxjRXNmUkdOOU5MVDJxZEc3S1dnODBJcmpB?oc=5","source":"search_cn_open_weight_labs","type":"news","summary_1line":"DeepSeek Open Sources Inference Tech to Lower AI Costs PYMNTS.com","sid":"b7d3f6df6ecbe2c9","published":"2026-10-20T18:55:08+00:00"},{"title":"DeepSeek open-sources DeepInfer, the inference framework behind its production services","url":"https://news.google.com/rss/articles/CBMiwAFBVV95cUxOSWRWVTdKcHNNZVpFRmJSNFZKVHdSQVpxbTNGT1AxLWJ4T1htaWlubWo0MTVBSXdxcGJIOVozQjhKNUI3TW11MFlYZnRvcmhLNEJWN0t4M190dFgyQV9BY0hFTmU4eXNnYzNJbXpycEVzOUNuRVY4T1BnYThPRWpJV0dYYUJ1RHVsUURPOTB0a0dtM0dnWTk0cURhMHNQYzdSdDNoblpHc3pzVWJ2ZnM2T3JJNDI3YzdwQ183MjZwM0Q?oc=5","source":"search_cn_open_weight_labs","type":"news","summary_1line":"DeepSeek open-sources DeepInfer, the inference framework behind its production services SiliconANGLE","sid":"8da24ee4e99be00d","published":"2026-10-20T23:11:00+00:00","editor_note":"Names DeepInfer as the engine behind production services"}]},{"date":"2026-10-21","items":[{"title":"DeepSeek Opens DeepInfer, Presses OpenAI, Anthropic [2026]","url":"https://news.google.com/rss/articles/CBMiggFBVV95cUxOUERNc0tRaVcxc2JFR2pHQU1JcHRkZk5ySTVXUHlsOFhxaE5mdGJjOTY5ek9HQTFNdzVzQmtpcFY5LTlEako5aHRPSnBXdmg1ZVZ0N0JpRUEyS1dma28xUF9IdEVxRnM1ZjB3eHpMSGhjb1dlc0JvYmtBYjJqQmsxWTN3?oc=5","source":"search_cn_open_weight_labs","type":"news","summary_1line":"DeepSeek Opens DeepInfer, Presses OpenAI, Anthropic [2026] shattered.io","sid":"ecdc2bd2da306c52","published":"2026-10-21T00:26:32+00:00"},{"title":"DeepSeek Open-Sources Framework Behind Its High-Margin AI Inference","url":"https://news.google.com/rss/articles/CBMipwFBVV95cUxOcWxaVjdURVh2U0tyVVZTOVdPRlVrVi1sWGVCZWVSNGZweTl1NjY2YnVEdE1SXzVncGNnaU1qQWp4NUFfLWpVM1NJcmZFSzJwOWJCSlZuVTRDOUFocXREV0g1dWlHa2Njbng4SUZtN0tLMkF3V1VHbkJGUjVmNDIzZEYxNEpaR1ZXZWxObDZFMEo0cFlDLUstTTlUdHdSSjZpbWJDN0JjOA?oc=5","source":"search_cn_open_weight_labs","type":"news","summary_1line":"DeepSeek Open-Sources Framework Behind Its High-Margin AI Inference The Information","sid":"249a6af86a1f4f73","published":"2026-10-21T04:02:00+00:00"},{"title":"[AINews] DeepSeek DeepInfer: Open Source SOTA Inference running 155T toks/day/cluster making $22m/day at ~80% margin for a 81 day payback","url":"https://latent.space/p/ainews-deepseek-deepinfer-open-source","source":"latent_space","type":"news","summary_1line":"An open source “bomb” drops from China","sid":"13bafceb1538461d","published":"2026-10-21T06:54:09+00:00","editor_note":"Adds the 155T tokens/day and revenue claims"},{"title":"DeepSeek open-sources DeepInfer, the engine serving 155T tokens a day","url":"https://news.google.com/rss/articles/CBMipAFBVV95cUxORmtUMnZPWTFnenFJQUlENWc2eWJ4RTJoTUFoNkM2ZC04dnBWdjNhZzMxVHdvR1hHRnoxUkFIVWRfX1hwbldqRWtxanQyczJuRlpOUTZjSHI1SUZXQ21BcktycXdtUDdoUU02ZjAwUmtGQkRQX0RwRllwazY3anN4amFLWnFEZlpIRjl0YUV0cTBYUUY1YmhYRWxXUVZzLWl4Smp5cA?oc=5","source":"search_cn_open_weight_labs","type":"news","summary_1line":"DeepSeek open-sources DeepInfer, the engine serving 155T tokens a day Dealroom","sid":"c928ee663a751a10","published":"2026-10-21T08:39:03+00:00"},{"title":"DeepSeek Open-Sources DeepInfer, Its Production Inference Engine","url":"https://news.google.com/rss/articles/CBMingFBVV95cUxNT1RCUmFqMFgzVEZyOExGbjgwZjU0Tk9MWHVGS0tuT3BpMFptQU5UYUp1aWhlVVFacUxkRklyYXd4SjVzRjRURVFBX2FSZXdseHFhdmVta25MUG5fYUkzMWFGcENucU1xMTFHTTd1clF3OE1yX3ZGRTNkODlzeUJKSkFTcjY3elRwRkJtMEc2dFhoM1p6OWp0NFVOWkg3UQ?oc=5","source":"search_cn_open_weight_labs","type":"news","summary_1line":"DeepSeek Open-Sources DeepInfer, Its Production Inference Engine Technology Org","sid":"5a29dcd699a504fa","published":"2026-10-21T09:34:00+00:00"},{"title":"VX-One: Typed, Calibrated Decision Heads for Real-Time Perimeter Triage - An empirical study of unknown-attack honesty, cascade gray zones, and calibration under drift","url":"http://arxiv.org/abs/2610.30090v1","source":"arxiv_cs_lg","type":"paper","summary_1line":"Perimeter triage faces a latency/semantics trade-off: signature engines are fast and format-safe but blind to unseen attacks, while generative models are semantic but slow and produce unparseable output. In late 2026...","why_it_matters":"Matches feed focus: eval.","sid":"a0784d9bff0dcc3c","published":"2026-10-21T17:39:12+00:00"},{"title":"Can LLMs Improve Their Own Harness? A Controlled Study on ARC-AGI-3","url":"http://arxiv.org/abs/2610.30499v1","source":"arxiv_agent_systems_research","type":"paper","summary_1line":"A generally intelligent system acquires new skills efficiently without requiring humans to redesign its software scaffold for every new task. Yet language models rarely act alone. A harness is a symbolic program that...","why_it_matters":"Matches feed focus: agent, harness, eval.","sid":"ff59ea23eb8cc0be","published":"2026-10-21T22:29:47+00:00"}]},{"date":"2026-10-22","items":[{"title":"Better Reasoning, Not Just Bigger Models: A Structure-Grounded Framework for Efficient and Reliable Inference","url":"http://arxiv.org/abs/2610.30611v1","source":"arxiv_agent_systems_research","type":"paper","summary_1line":"Scaling large language models has substantially improved linguistic fluency, yet reliable autonomous agents require more than fluent generation. In multi-turn inference, current agentic systems often produce semantica...","why_it_matters":"Matches feed focus: agentic.","sid":"b3f4301ff8d83589","published":"2026-10-22T02:20:21+00:00"},{"title":"A framework for evaluating agentic AI systems","url":"https://deepsense.ai/blog/your-agent-passed-the-eval-should-you-ship-it","source":"hackernews_ai","type":"news","summary_1line":"A framework for evaluating agentic AI systems","why_it_matters":"Matches feed focus: agentic, eval.","sid":"520159e6ebe00b0f","published":"2026-10-22T07:32:51+00:00"},{"title":"Tracing Hallucinations in LLM Reasoning for UML Class Diagram Generation: An Empirical Taxonomy and Mitigation Study","url":"http://arxiv.org/abs/2610.30946v1","source":"arxiv_llm_reliability","type":"paper","summary_1line":"Large Language Models (LLMs) can generate UML class diagrams from natural language requirements, but hallucinations may compromise the correctness and completeness of the resulting diagrams. Existing evaluations mostl...","why_it_matters":"Matches feed focus: evaluation.","sid":"60ee2bd6a79177f2","published":"2026-10-22T08:59:27+00:00"},{"title":"How Well Do Decision Models Transfer to Vietnamese? An Empirical Study of Text Classification","url":"http://arxiv.org/abs/2610.31422v1","source":"arxiv_cs_cl","type":"paper","summary_1line":"Recent advances in specialized decision models have expanded the possibilities for zero-shot classification, yet their effectiveness on Vietnamese remains underexplored. This study evaluates four specialized decision...","why_it_matters":"Matches feed focus: evaluation.","sid":"d36c96c2a6372e9b","published":"2026-10-22T15:24:16+00:00"},{"title":"An Empirical Study of Recent Large Language Models for Knowledge Graph to Text Generation on Standard Benchmarks","url":"http://arxiv.org/abs/2610.31529v1","source":"arxiv_cs_cl","type":"paper","summary_1line":"Knowledge graph-to-text (KG-to-text) generation verbalizes a set of knowledge graph triples into fluent natural language. While fine-tuned encoder-decoder models have long been the standard approach, recent LLMs are b...","why_it_matters":"Matches feed focus: evaluation.","sid":"c888971c48a268c3","published":"2026-10-22T16:36:35+00:00"},{"title":"Social-HUMINT: An LLM-Augmented Unified Framework for Measuring Social Media Exposure","url":"http://arxiv.org/abs/2610.31573v1","source":"arxiv_cs_ai","type":"paper","summary_1line":"In the post-API landscape, social-media profiling has become inexpensive and automated: Large Language Models (LLMs) can extract structured information from unstructured content and deduce attributes users never discl...","why_it_matters":"Matches feed focus: agent, eval.","sid":"4d53557cac2f8289","published":"2026-10-22T17:13:46+00:00"}]},{"date":"2026-10-23","items":[{"title":"Do Language Models Trust Their Own Justifications? A Study on Functional Consistency","url":"http://arxiv.org/abs/2610.32665v1","source":"arxiv_cs_ai","type":"paper","summary_1line":"Large Language Models (LLMs) have been widely adopted in text classification tasks, where they not only output class predictions but also generate explanations that highlight the tokens deemed most relevant for reachi...","why_it_matters":"Matches feed focus: evaluation.","sid":"72b88f8041b34aeb","published":"2026-10-23T14:08:32+00:00"}]},{"date":"2026-10-24","items":[{"title":"Why founders keep failing to vibecode their own LLM inference engine","url":"https://news.google.com/rss/articles/CBMinAFBVV95cUxNWFQyRXJtQU5CWDVfWjZfUXpLaDFFUkFJc2tkWWE3RDljeVRhRmoyRXhIS1A1cG5Hd0VxTDhZdlNENmY1czlEQ19uZjVKOXRja1ZMeW5DNXpNT1hydjNKUlNLaFpZNlV2OHI5NkZIaGJFbUs0OUpMa1JqLUc2WHdjN2VmTHY0eDJCZDI2aW5mQVFudnl2RXdWSDJWNzE?oc=5","source":"search_llm_ops_news","type":"news","summary_1line":"Why founders keep failing to vibecode their own LLM inference engine Startup Fortune","sid":"11667337918f4f3c","published":"2026-10-24T00:24:08+00:00"}]},{"date":"2026-10-25","items":[{"title":"EvolveMAS: A Unified Self-Evolution Framework for Decentralized Multi-Agent Systems","url":"http://arxiv.org/abs/2610.34450v1","source":"arxiv_agent_systems_research","type":"paper","summary_1line":"Decentralized multi-agent systems (MAS) enable large language model agents to interact without centralized control. Agents need to adapt to other agents at test time and reuse accumulated experience to refine their in...","why_it_matters":"Matches feed focus: agent, evaluation.","sid":"378efdb4d515dcd8","published":"2026-10-25T17:09:49+00:00"}]},{"date":"2026-10-26","items":[{"title":"CEPO: A General Framework for Improving Prompt Optimization in Multi-Agent Systems","url":"http://arxiv.org/abs/2610.34722v1","source":"arxiv_agent_systems_research","type":"paper","summary_1line":"Multi-agent systems (MAS) require both effective coordination and specialized task execution. Existing prompt optimization approaches typically couple these functions within agent-specific prompts. We introduce Coordi...","why_it_matters":"Matches feed focus: agent.","sid":"7c1fed0d11e9b516","published":"2026-10-26T02:19:53+00:00"}]},{"date":"2026-10-29","items":[{"title":"A Unified Hessian-Guided Framework for Post-Training Quantization","url":"http://arxiv.org/abs/2610.40345v1","source":"arxiv_cs_lg","type":"paper","summary_1line":"Low-bit quantization enables efficient deployment of large language models by reducing memory footprint and computational cost. Achieving high accuracy at low precision requires selecting quantization ranges, linear transformations, and weight rounding decisions. Existing approac","sid":"05c46a37b489922f","published":"2026-10-29T16:06:01+00:00"}]},{"date":"2026-10-30","items":[{"title":"Confident Fools or Careful Searchers? A Behavioural Study of When LLMs Search Beyond Their Static Internal Memory","url":"http://arxiv.org/abs/2610.41124v1","source":"arxiv_cs_ai","type":"paper","summary_1line":"Agentic LLM assistants must repeatedly decide whether to answer from memory or to call a tool such as web search. Unnecessary searches cost latency and money; missing ones yield confident but unsupported answers. Prio...","why_it_matters":"Matches feed focus: agentic, eval.","sid":"0b386b33c8983aa1","published":"2026-10-30T01:28:38+00:00"},{"title":"Compression Technique Interactions in Small Language Models: A Factorial Ablation Study","url":"http://arxiv.org/abs/2611.00464v1","source":"arxiv_cs_ai","type":"paper","summary_1line":"Deploying small language models on resource-constrained hardware requires compression, yet practitioners lack systematic guidance on how quantization, pruning, and post-compression recovery interact when combined. We...","why_it_matters":"Matches feed focus: evaluation.","sid":"b80471ec3bc792a2","published":"2026-10-30T23:36:03+00:00"}]},{"date":"2026-10-31","items":[{"title":"WeSearch: A Unified Model-Centric Retrieval Framework for WeChat Search","url":"http://arxiv.org/abs/2611.00530v1","source":"arxiv_cs_cl","type":"paper","summary_1line":"We present WeSearch, a unified, model-centric retrieval framework for large-scale, continuously evolving content search. Prevailing methods assign documents fixed hierarchical Semantic IDs, implicitly assuming each do...","why_it_matters":"Matches feed focus: eval.","sid":"96207d6b2522935e","published":"2026-10-31T01:35:20+00:00"},{"title":"Inference engine in game scripting language using macros","url":"https://dasllama.io/articles/2026-10-28-the-need-for-speed.html","source":"hackernews_ai","type":"news","summary_1line":"Inference engine in game scripting language using macros","sid":"c7641f54401d6258","published":"2026-10-31T19:17:31+00:00"},{"title":"DeepSeek put the architecture for their always-on autonomous agent on GitHub","url":"https://github.com/deepseek-ai/dsh-superagent-prototype","source":"hackernews_ai","type":"news","summary_1line":"DeepSeek put the architecture for their always-on autonomous agent on GitHub","why_it_matters":"Matches feed focus: agent.","sid":"5bd78febd4a59e34","published":"2026-10-31T23:31:26+00:00","editor_note":"First report of the always-on agent architecture on GitHub"}]},{"date":"2026-11-01","items":[{"title":"DeepSeek Open Sources the Architecture Behind Its Super-Agent","url":"https://news.google.com/rss/articles/CBMikwFBVV95cUxPTFhOWk9yYkRyNkE0d3FxeWNtVDdmcGFNTHBxeWxINk1iUGJMQTd4c2lXRkNMY3NjejVyTGIwZmdYQXUweTdNUHAwblI5M2dGNmd5TWR5SXo4Rmx6d3ZCWjZhX21YZHNoMDNhT3cycEtxd2FGeEdPbkR3UWRiRkNwbTJyandOcEtQXzRyUkxhdlhybTg?oc=5","source":"search_cn_open_weight_labs","type":"news","summary_1line":"DeepSeek Open Sources the Architecture Behind Its Super-Agent Startup Fortune","why_it_matters":"Matches feed focus: agent.","sid":"7ae79fa7e63f017f","published":"2026-11-01T00:18:27+00:00"},{"title":"DeepSeek Harness Team Open-Sources a Super-agent Prototype That Finds and Fixes Bugs on Its Own","url":"https://news.google.com/rss/articles/CBMisAFBVV95cUxOaUpld3paRUZYenVQaE96RThTb1p3QXUyNm5OSWVtNnZoRXVQdFJPX2VyYVJxNzFqYXliNk1pQUJETkdQdzI2Qm40TDBkTDhSTmZQX0RKYTZvSjRyRFJXWU5QdjVnOHJZV3hlaUgwUmZ1aXg4V3kxU002QnBwbngteU5hckdXM0dKOFlqMlpIZXNzMG5fNHVuTm8zTWlHdktMeEM1MzU1Z3owcW9pMm1wZA?oc=5","source":"search_cn_open_weight_labs","type":"news","summary_1line":"DeepSeek Harness Team Open-Sources a Super-agent Prototype That Finds and Fixes Bugs on Its Own Pandaily","why_it_matters":"Matches feed focus: agent, harness.","sid":"7f903ef14a1441d8","published":"2026-11-01T01:53:20+00:00"},{"title":"DeepSeek Unveils Open-Source 'Super-Agent' Coding Tool in Direct Challenge to Anthropic's Claude Code","url":"https://news.google.com/rss/articles/CBMidkFVX3lxTE04Z3JfR3VDQUZIMjhMdXZQeWltTmxyYmk1SVNqM3ExYUNZaVVpTzNYelZGaXI5Z0MwbEtrRmhRS0xSblJpMmQtZm12S1V3b1ZzYlFSckxTN1ppT1Y1U1lQWElMVDF3NFBsb0ozMGhLNmw4ZGFnZFE?oc=5","source":"search_cn_open_weight_labs","type":"news","summary_1line":"DeepSeek Unveils Open-Source 'Super-Agent' Coding Tool in Direct Challenge to Anthropic's Claude Code BigGo Finance","why_it_matters":"Matches feed focus: agent, claude code.","sid":"d795ff607b219d9a","published":"2026-11-01T06:25:00+00:00"},{"title":"DeepSeek Open-Sources dsh-superagent, a Router-Worker Coding Agent","url":"https://news.google.com/rss/articles/CBMilgFBVV95cUxOVGVmcmxZdFpKS25HOEFNbnFrZmRkd0ZRVFNjUlNMMEEwQXZ3YjFSTm9nNGU3NGZMeFBQWEhtZUlvQU9Va3UzeDlIZXVMaW5QM0w1YUpscHUtMHVqeDlSTnotOTZLNXNUV09abkc3U0VGZjU0MTlLaWJJdGtEWnhTMFlkQWJuWWFVUmRnTFB4eXlUbDNiY0HSAZYBQVVfeXFMTlRlZnJsWXRaSktuRzhBTW5xa2ZkZHdGUVRTY1JTTDBBMEF2d2IxUk5vZzRlNzRmTHhQUFhIbWVJb0FPVWt1M3g5SGV1TGluUDNMNWFKbHB1LTB1ang5Uk56LTk2SzVzVFdPWm5HN1NFRmY1NDE5S2liSXRrRFp4UzBZZEFibllhVVJkZ0xQeHl5VGwzYmNB?oc=5","source":"search_agent_engineering_news","type":"news","summary_1line":"DeepSeek Open-Sources dsh-superagent, a Router-Worker Coding Agent MarkTechPost","why_it_matters":"Matches feed focus: agent.","sid":"8ffb3cb5b711c694","published":"2026-11-01T07:47:17+00:00","editor_note":"Names dsh-superagent and its router-worker design"}]},{"date":"2026-11-02","items":[{"title":"Flare, Nvidia's DGX Spark Inference Engine","url":"https://github.com/NVIDIA/spark-local-inference","source":"hackernews_ai","type":"news","summary_1line":"Flare, Nvidia's DGX Spark Inference Engine","sid":"e6077e10f16ce2d8","published":"2026-11-02T01:11:27+00:00"},{"title":"DeepSeek open-sources dsh-superagent research prototype","url":"https://news.google.com/rss/articles/CBMihAFBVV95cUxQcFMxa1Q5RUs4eUlwcWV2bWdMNU1NR1gwbXpGUnFwZThweTRTU1VGQUZKTkh5Ukp1NnJPMG5FdkZCbXZSeDQwWXJvNmVON09JMzVvaUlDVTc4RjJpakxSdDF5aUJVSkdIbzBpajdyM0steEp6NUpDWmk3c0t3Zk5yXzd1V3o?oc=5","source":"search_cn_open_weight_labs","type":"news","summary_1line":"DeepSeek open-sources dsh-superagent research prototype TNGlobal","why_it_matters":"Matches feed focus: agent.","sid":"a8f5be6d64b1c57f","published":"2026-11-02T02:27:13+00:00"},{"title":"DeepSeek team open-sources proactive coding agent research prototype","url":"https://news.google.com/rss/articles/CBMiowFBVV95cUxPZlhRZXhnYlJmX1RiOUdNamRqUXJUU2NycHNuZWtGeEVUejJxNVdEQkQzUU1HMkxHR1dXeDhBVDlPRmVyUHFoMU96VkdBYzlRUFRSWkFFTGVBWWFjNXRRaUlld2d2alpFdVd0VXFlR2tMeXNJUWdDc3c1cG5SR3BPRVd0OVRFLXBvMWhJZ0hzRG1qU2VQazBDQjNQNHl2cXRJZFZV?oc=5","source":"search_cn_open_weight_labs","type":"news","summary_1line":"DeepSeek team open-sources proactive coding agent research prototype TechNode","why_it_matters":"Matches feed focus: agent.","sid":"4141e315102a7089","published":"2026-11-02T04:32:41+00:00"},{"title":"DeepSeek open-sources Super-Agent, a coding agent that watches team chats for work","url":"https://news.google.com/rss/articles/CBMidkFVX3lxTE5nREY5eGxfa2tXV0ZxeHZpekdIaS05UWdfYUxVTHNIR01TdnU0MEhxaHA2NE1iSHE2MnpHRUgtM3VvR2x2ZEpKQ1ljYVFEME9lQTEtOHNXSHdkT0Jjb0VPTUgtYnNVdEQ2RXBlMERLOTRTcF9ic2c?oc=5","source":"search_cn_open_weight_labs","type":"news","summary_1line":"DeepSeek open-sources Super-Agent, a coding agent that watches team chats for work Help Net Security","why_it_matters":"Matches feed focus: agent.","sid":"faef8337e561998f","published":"2026-11-02T07:09:55+00:00"},{"title":"An Axiomatic Framework for Reinforcement Learning in Open Multi-Agent Systems","url":"http://arxiv.org/abs/2611.03301v1","source":"arxiv_agent_systems_research","type":"paper","summary_1line":"Open multi-agent systems, where only a subset of agents are controllable, are central to real-world domains such as smart grids, yet lack a principled foundation for reinforcement learning algorithm design. We study this setting through n-agent ad hoc teamwork (NAHT), a recent fr","why_it_matters":"Matches feed focus: agent.","sid":"bce25dc1c825a3e7","published":"2026-11-02T15:36:22+00:00"},{"title":"A Cross-Domain Matrix Engine for Signal Processing and Deep Neural Net Inference","url":"http://arxiv.org/abs/2611.03503v1","source":"arxiv_cs_ai","type":"paper","summary_1line":"Matrix multiplication dominates both PHY signal processing and Deep Neural Network (DNN) inference, yet these domains place opposing demands on the matrix engine. Signal processing requires low-latency execution of relatively small matrix operations, whereas DNN inference, partic","sid":"fc7c1780406958b5","published":"2026-11-02T17:40:14+00:00"}]},{"date":"2026-11-03","items":[{"title":"Is hosting an open-source inference engine profitable?","url":"https://userecurrence.ai/research/is-hosting-open-source-inference-profitable","source":"hackernews_ai","type":"news","summary_1line":"Is hosting an open-source inference engine profitable?","sid":"2f11643a689186b4","published":"2026-11-03T04:27:10+00:00"},{"title":"Aligning Language Models with General Preferences: An Empirical Study of Direct Nash Optimization","url":"http://arxiv.org/abs/2611.05373v1","source":"arxiv_cs_cl","type":"paper","summary_1line":"Direct Nash Optimization (DNO) aligns a language model to a general preference function by approximating the Nash equilibrium of a two-player preference game. Its practical variant, DNO-Prct, iterates on-policy sampling, LLM-based rating of the samples together with a teacher's g","why_it_matters":"Matches feed focus: evaluation.","sid":"5ab1b6ecf4ca3e7a","published":"2026-11-03T18:59:15+00:00"}]},{"date":"2026-11-04","items":[{"title":"DeepSeek Open-Sources Super-Agent, a Coding Agent That Works From Team Chat","url":"https://news.google.com/rss/articles/CBMioAFBVV95cUxPOWRQWmRtMFZFaDVDWGpNYXg2WV9pcVlKcVItSXlrS0twTU55X3VPZU1OSFFOUGhQUmlMcXg5dkVIemtCMFYydFdxc2lDOTQzUFZYNjhZOEpBTHlzdmt2V0JRRXlqRG9DZ0xnSFI3Y0stYnNyRFZpNzg5ekh6czVoVW5vUk40WEE5ZVA1ZVNseXplX3Z3b2hGTUFiN1RvS3Vk?oc=5","source":"search_cn_open_weight_labs","type":"news","summary_1line":"DeepSeek Open-Sources Super-Agent, a Coding Agent That Works From Team Chat Technology Org","why_it_matters":"Matches feed focus: agent.","sid":"28c838a66808367f","published":"2026-11-04T09:57:00+00:00","editor_note":"Latest DeepSeek coverage of the team-chat workflow"}]},{"date":"2026-11-06","items":[{"title":"Supercharging XLA-Based Inference: Announcing the Launch of XLA Serving Engine","url":"https://opensource.googleblog.com/2026/11/supercharging-xla-inference-xla-serving-engine.html","source":"hackernews_ai","type":"news","summary_1line":"Supercharging XLA-Based Inference: Announcing the Launch of XLA Serving Engine","sid":"9b6ce22e2f29b264","published":"2026-11-06T15:21:03+00:00","editor_note":"Newest item: a separate XLA-based serving engine"}]}],"editorial":{"tldr":"DeepSeek tied its V4.1-Flash model to its own agent harness on Oct 17, then on Oct 20 open-sourced DeepInfer, the inference engine behind its production services. On Oct 31 it published the architecture of an always-on coding agent, dsh-superagent.","stale":false,"whats_new":"Coverage of the dsh-superagent prototype, a router-worker coding agent that watches team chat for work, continued through Nov 4, while the rest of the thread is third-party inference-engine news.","why_it_matters":"DeepSeek is releasing both the serving layer (DeepInfer, reportedly 155T tokens/day) and an agent harness, so platform teams can inspect a production-scale stack, but the cost and throughput figures come from DeepSeek and press reports, not independent benchmarks.","take_for_builders":"Before adopting DeepInfer or dsh-superagent, check license terms and hardware support, and benchmark them against your current serving stack rather than relying on the quoted token-per-day numbers.","status":{"state":"Developing","tone":"rising","changed":"2026-11-04","detail":"Two DeepSeek open-source drops, DeepInfer and dsh-superagent, with no independent benchmarks yet."},"beats":[{"kicker":"HARNESS","tone":"launch","headline":"V4.1-Flash-1017 tied tighter to DeepSeek's own harness","sids":["ce1f00045df6fba8"]},{"kicker":"DEEPINFER","tone":"turn","headline":"DeepSeek open-sources DeepInfer, its production inference engine","summary":"Wide press pickup of the engine, with claims of 155T tokens a day per cluster.","sids":["6ca9e1ceeea5a770","be61663792fa0af3","861b6347e652747b","b7d3f6df6ecbe2c9","8da24ee4e99be00d","ecdc2bd2da306c52","249a6af86a1f4f73","13bafceb1538461d","c928ee663a751a10","5a29dcd699a504fa"]},{"kicker":"ADJACENT","tone":"neutral","headline":"Loosely related research and framework papers clustered in","summary":"Mostly arXiv and discussion items that share keywords with the thread rather than reporting on DeepSeek.","sids":["a0784d9bff0dcc3c","ff59ea23eb8cc0be","b3f4301ff8d83589","520159e6ebe00b0f","60ee2bd6a79177f2","d36c96c2a6372e9b","c888971c48a268c3","4d53557cac2f8289","72b88f8041b34aeb","378efdb4d515dcd8","7c1fed0d11e9b516","05c46a37b489922f","0b386b33c8983aa1","b80471ec3bc792a2","96207d6b2522935e","bce25dc1c825a3e7","5ab1b6ecf4ca3e7a"]},{"kicker":"SUPER-AGENT","tone":"now","headline":"dsh-superagent: a router-worker coding agent goes open source","summary":"Positioned against Claude Code; works from team chat.","sids":["5bd78febd4a59e34","7ae79fa7e63f017f","7f903ef14a1441d8","d795ff607b219d9a","8ffb3cb5b711c694","a8f5be6d64b1c57f","4141e315102a7089","faef8337e561998f","28c838a66808367f"]},{"kicker":"ECOSYSTEM","tone":"neutral","headline":"Other inference engines and serving economics in the same window","summary":"llm-d 1.0, Nvidia Flare, XLA Serving Engine and hosting-profitability discussion.","sids":["db7388ca69770009","0714a9511cb60f66","11667337918f4f3c","c7641f54401d6258","e6077e10f16ce2d8","fc7c1780406958b5","2f11643a689186b4","9b6ce22e2f29b264"]}],"open_questions":["Will anyone publish independent throughput or cost benchmarks of DeepInfer outside DeepSeek's own figures?","Which models and hardware does DeepInfer support beyond DeepSeek's own?","Is dsh-superagent usable beyond a research prototype, and under what license?"],"generated_at":"2026-11-07T00:12:01.132749+00:00"}}