Research by Aamish Ahmad Beg
Mastra’s Memory is SotA on BEAM
Oct 29, 2026 · 25 min read
Observational Memory has the top published scores on BEAM at 100K (83.3%) and 1M tokens (80.6%), with default settings and no benchmark tuning.
Decision Models Get Context Rot Too
Oct 22, 2026 · 10 min read
We padded 400 questions with up to 26k tokens and compared Jev, Clef and GPT-6 Luna Decisions with reasoning and non-reasoning LLMs.