AI Research & Science
Peer-reviewed breakthroughs, university studies, and lab discoveries — explained in plain English. AI Maestro tracks the frontiers of machine learning research, neuroscience meets AI, and the science driving the next wave of intelligent systems.
Building the foundation for an autonomous enterprise
Woodside Energy in Western Australia has been using artificial intelligence since 2015. The company has spent a decade building predictive models and…
Top stories
LeRobot v0.6.0: Imagine, Evaluate, Improve
2d ago
AI search agents don’t fail at searching, they fail at asking the right questions when queries get ambiguous
3d ago
Anthropic Launches Claude Science Beta: A Multi-Agent AI Workbench for Reproducible Genomics, Proteomics, and Cheminformatics Pipelines
3d ago
NVIDIA HORIZON: A Hands-Free Agent that Evolves Git Worktrees and Hits 100% RTL Benchmark Completion
3d agoMore ai research & science
Benchmarking the new b9200 update: Optimizing Qwen 3.6 27B mtp for Hermes Agent on a single RTX 3090
“`html Benchmarking the new b9200 update: Optimizing Qwen 3.6 27B mtp for Hermes Agent…
18 May 2026
Benchmarking vLLM vs SGLang vs llama.cpp on a mixed Blackwell/Ada cluster
Benchmarking vLLM vs SGLang vs llama.cpp on a mixed Blackwell/Ada cluster Summary of Benchmarks…
18 May 2026
Seeking local LLM advice for cybersecurity work.
“`html A British AI enthusiast is seeking advice on using a local Large Language…
17 May 2026
A Coding Implementation to Compress and Benchmark Instruction-Tuned LLMs with FP8, GPTQ, and SmoothQuant Quantization using llmcompressor
Overview In this tutorial, we explore how to apply post-training quantization to an instruction-tuned…
17 May 2026
Qwen3.6-27B MTP depth benchmark, RTX 3090Ti
“`html A new benchmark for the Qwen 3.6-27B model has been conducted using an…
17 May 2026
85 GPU-hours comparing 5 abliteration methods on Qwen3.6-27B: benchmarks, safety, weight forensics – Abliterlitics
The six models Name Type Base Qwen/Qwen3.6-27B Heretic llmfan46/Qwen3.6-27B-uncensored-heretic-v2 HauhauCS HauhauCS/Qwen3.6-27B-Uncensored-HauhauCS-Aggressive Huihui huihui-ai/Huihui-Qwen3.6-27B-abliterated AEON…
17 May 2026
Qwen 3.6-27B Dense with MTP on Strix Halo Windows – Benchmarks
**Editorial Brief** A new benchmark study has been released for Qwen 3.6-27B, a large…
17 May 2026
New math benchmark reveals AI models confidently solve problems that have no solution
“`html A new AI benchmark called SOOHAK has been created by a consortium of…
17 May 2026
Research repository ArXiv will ban authors for a year if they let AI do all the work
ArXiv Takes Action Against Unchecked AI-Generated Papers Key Takeaways ArXiv, a major repository for…
16 May 2026