#openweightmodel — Public Fediverse posts
Live and recent posts from across the Fediverse tagged #openweightmodel, aggregated by home.social.
-
DeepSeek-V4: Low-Cost Logic via Hybrid Attention Architectures Presentational View Introduction There is an evident inclination toward novel innovations that incorporate unique structural modificat...
#open-weight-model #agentic-search #deepseek-v4 #hybrid-attention #coding-agents
Origin | Interest | Match -
Mô hình trọng lượng mở tốt cho việc sử dụng công cụ là gì? #ToolUse #AgenticUseCases #MôHìnhTrọngLượngMở #CôngCụ #AI #TríTuệNhânTạo #OpenWeightModel #ArtificialIntelligence
https://www.reddit.com/r/LocalLLaMA/comments/1owljcm/good_open_weight_model_for_tool_use/
-
#DeepSeek’s #R1 #AImodel, a cheaper rival to US tools, excels at reasoning tasks and is the most popular #openweightmodel on Hugging Face. The model, trained for US$294,000 using Nvidia’s H800 chips, employs pure reinforcement learning, rewarding correct answers rather than following human examples. This approach, detailed in a peer-reviewed Nature paper, has influenced AI research in 2025. https://www.nature.com/articles/d41586-025-03015-6?eicker.news #tech #media #news
-
#DeepSeek’s #R1 #AImodel, a cheaper rival to US tools, excels at reasoning tasks and is the most popular #openweightmodel on Hugging Face. The model, trained for US$294,000 using Nvidia’s H800 chips, employs pure reinforcement learning, rewarding correct answers rather than following human examples. This approach, detailed in a peer-reviewed Nature paper, has influenced AI research in 2025. https://www.nature.com/articles/d41586-025-03015-6?eicker.news #tech #media #news
-
#DeepSeek’s #R1 #AImodel, a cheaper rival to US tools, excels at reasoning tasks and is the most popular #openweightmodel on Hugging Face. The model, trained for US$294,000 using Nvidia’s H800 chips, employs pure reinforcement learning, rewarding correct answers rather than following human examples. This approach, detailed in a peer-reviewed Nature paper, has influenced AI research in 2025. https://www.nature.com/articles/d41586-025-03015-6?eicker.news #tech #media #news
-
#DeepSeek’s #R1 #AImodel, a cheaper rival to US tools, excels at reasoning tasks and is the most popular #openweightmodel on Hugging Face. The model, trained for US$294,000 using Nvidia’s H800 chips, employs pure reinforcement learning, rewarding correct answers rather than following human examples. This approach, detailed in a peer-reviewed Nature paper, has influenced AI research in 2025. https://www.nature.com/articles/d41586-025-03015-6?eicker.news #tech #media #news
-
#DeepSeek’s #R1 #AImodel, a cheaper rival to US tools, excels at reasoning tasks and is the most popular #openweightmodel on Hugging Face. The model, trained for US$294,000 using Nvidia’s H800 chips, employs pure reinforcement learning, rewarding correct answers rather than following human examples. This approach, detailed in a peer-reviewed Nature paper, has influenced AI research in 2025. https://www.nature.com/articles/d41586-025-03015-6?eicker.news #tech #media #news
-
LG AI Research Institute has unveiled EXAONE 4.0, South Korea's first hybrid AI model, surpassing global benchmarks and expanding commercial API access.
#YonhapInfomax #LG #EXAONE40 #HybridAI #OpenWeightModel #HuggingFace #Economics #FinancialMarkets #Banking #Securities #Bonds #StockMarket
https://en.infomaxai.com/news/articleView.html?idxno=72077 -
LG AI Research Institute has unveiled EXAONE 4.0, South Korea's first hybrid AI model, surpassing global benchmarks and expanding commercial API access.
#YonhapInfomax #LG #EXAONE40 #HybridAI #OpenWeightModel #HuggingFace #Economics #FinancialMarkets #Banking #Securities #Bonds #StockMarket
https://en.infomaxai.com/news/articleView.html?idxno=72077 -
LG AI Research Institute has unveiled EXAONE 4.0, South Korea's first hybrid AI model, surpassing global benchmarks and expanding commercial API access.
#YonhapInfomax #LG #EXAONE40 #HybridAI #OpenWeightModel #HuggingFace #Economics #FinancialMarkets #Banking #Securities #Bonds #StockMarket
https://en.infomaxai.com/news/articleView.html?idxno=72077 -
LG AI Research Institute has unveiled EXAONE 4.0, South Korea's first hybrid AI model, surpassing global benchmarks and expanding commercial API access.
#YonhapInfomax #LG #EXAONE40 #HybridAI #OpenWeightModel #HuggingFace #Economics #FinancialMarkets #Banking #Securities #Bonds #StockMarket
https://en.infomaxai.com/news/articleView.html?idxno=72077