DeepSeek-v4-Pro-Max
PulseAugur coverage of DeepSeek-v4-Pro-Max — every cluster mentioning DeepSeek-v4-Pro-Max across labs, papers, and developer communities, ranked by signal.
-
Poolside's Laguna S 2.1 open model surpasses DeepSeek-V4-Pro-Max
Poolside has released Laguna S 2.1, an 118-billion parameter Mixture-of-Experts model that outperforms the 1.6-trillion parameter DeepSeek-V4-Pro-Max on the Terminal-Bench 2.1 benchmark. This American-made open model bo…
-
Open Source LLMs Explode: DeepSeek V4-Pro-Max Hits 1.6T Parameters
April 2026 marked a significant surge in open-source large language models, with DeepSeek releasing its V4-Pro-Max model boasting 1,600 billion parameters. Four other models also surpassed the trillion-parameter thresho…
-
New 'Misattribution Gap' Attack Targets AI Memory Layers
A new research paper, "The Misattribution Gap," introduces "Semantic Norm Drift" (SND) as a novel attack vector for agentic AI systems. This attack exploits the memory layer, making it difficult to distinguish from mode…