Ingenix Says Thalamus Model Ranks First on LongevityBench

Warsaw-based AI company Ingenix says its 9-billion-parameter Thalamus Reasoning model ranked first among 29 models on LongevityBench, a benchmark for language models in cellular aging research. The company says it is open-sourcing a version called Thalamus Decision and is developing a 95-billion-parameter model. (e22ff8e7-de13-4070-8cf0-3a9a4f6c9671; cba67dd4-6c31-45d4-9b5b-f054ac089170)

• Ingenix says Thalamus Reasoning placed first across 14 ranked tasks, with a mean rank of 1.42, and led on 11 tasks. (e22ff8e7-de13-4070-8cf0-3a9a4f6c9671; cba67dd4-6c31-45d4-9b5b-f054ac089170)· 2 sources
• The company says it will release Thalamus Decision as open source for academic and commercial researchers; it also says it is developing a 95-billion-parameter model. (e22ff8e7-de13-4070-8cf0-3a9a4f6c9671; cba67dd4-6c31-45d4-9b5b-f054ac089170)· 2 sources
• Ingenix reports a DNA methylation age error of 3.36 years and transcriptomics ordering accuracy of 0.828; it says the benchmark data was not used to train Thalamus Reasoning. (e22ff8e7-de13-4070-8cf0-3a9a4f6c9671; cba67dd4-6c31-45d4-9b5b-f054ac089170)· 2 sources