https://ai-news.ypetit.net/sources/amazon-science/ Latest publications from Amazon Science, aggregated by AI News from the official source. en Fri, 23 Oct 2026 19:08:30 GMT https://www.amazon.science/blog/better-speech-synthesis-for-audiobookshttps://www.amazon.science/blog/better-speech-synthesis-for-audiobooksFri, 23 Oct 2026 15:20:02 GMTresearch“Controllable text to speech” makes speech more expressive, character voices more distinct and consistent. https://www.amazon.science/blog/the-science-behind-torchneuron-amazons-new-pytorch-backend-for-aws-trainium-chipshttps://www.amazon.science/blog/the-science-behind-torchneuron-amazons-new-pytorch-backend-for-aws-trainium-chipsThu, 22 Oct 2026 19:36:42 GMTresearchRuntime optimizations and a thorough evaluation process driven by agentic AI help ensure that models written in PyTorch will execute efficiently and reliably on Trainium systems. https://www.amazon.science/blog/strengthening-ai-safety-evaluation-through-independent-third-party-reviewhttps://www.amazon.science/blog/strengthening-ai-safety-evaluation-through-independent-third-party-reviewTue, 20 Oct 2026 17:57:55 GMTresearchHow Amazon partnered with AVERI to independently assess — and improve — our internal safety testing of Amazon models. https://www.amazon.science/blog/can-we-change-an-ai-models-architecture-without-retraining-it-from-scratchhttps://www.amazon.science/blog/can-we-change-an-ai-models-architecture-without-retraining-it-from-scratchMon, 12 Oct 2026 17:36:18 GMTresearch“Priming” reuses pretrained Transformer weights to build models with more efficient architectures at less than 1% of the original training cost and with throughputs up to twice as high. https://www.amazon.science/blog/graph-centric-agentic-intelligencehttps://www.amazon.science/blog/graph-centric-agentic-intelligenceThu, 01 Oct 2026 12:54:31 GMTresearchAugmenting a network graph with agentic AI produces a “digital twin” that can help isolate network failures. https://www.amazon.science/blog/a-kernel-centric-path-to-real-time-video-generation-on-trainiumhttps://www.amazon.science/blog/a-kernel-centric-path-to-real-time-video-generation-on-trainiumFri, 25 Sep 2026 14:55:00 GMTresearchUsing the Neuron Kernel Interface, a Reactor–AWS collaboration tackled the dynamic shapes, memory access patterns, and cache management that make real-time autoregressive diffusion hard—building techniques that generalize across models. https://www.amazon.science/news/amazon-launches-research-initiative-with-stanford-university-to-advance-ai-and-sciencehttps://www.amazon.science/news/amazon-launches-research-initiative-with-stanford-university-to-advance-ai-and-scienceMon, 21 Sep 2026 19:31:59 GMTresearchThe collaboration aims to advance research while broadening participation and translating discovery into real-world solutions. https://www.amazon.science/blog/advancing-ai-for-biology-teaching-models-to-design-and-characterize-antibodieshttps://www.amazon.science/blog/advancing-ai-for-biology-teaching-models-to-design-and-characterize-antibodiesMon, 21 Sep 2026 17:51:00 GMTresearchThree new papers from Amazon Bio Discovery address bottlenecks in AI-driven antibody engineering, from benchmarking binding predictors to experimentally validating de novo design. https://www.amazon.science/blog/why-dont-machine-learning-research-agents-overfithttps://www.amazon.science/blog/why-dont-machine-learning-research-agents-overfitThu, 10 Sep 2026 15:03:39 GMTresearchNew research indicates that AI agents learn compressible models of data, which don’t have enough space to enable memorization. https://www.amazon.science/blog/developing-provably-correct-rust-code-with-verushttps://www.amazon.science/blog/developing-provably-correct-rust-code-with-verusMon, 31 Aug 2026 15:35:33 GMTresearchHow the Verus "program verifier", which automatically checks code against a mathematical specification of its functionality, helps increase security assurance in software projects. https://www.amazon.science/blog/when-llm-judges-agree-should-we-believe-themhttps://www.amazon.science/blog/when-llm-judges-agree-should-we-believe-themWed, 26 Aug 2026 17:10:40 GMTresearchDiscounting the opinions of LLM judges with highly correlated outputs ensures that panels of judges reflect a true diversity of perspectives.