Customer-obsessed science
Research areas
-
November 16, 202611 min readNew open-source tool turns existing test suites into sets of formal-verification tasks, which catch buggy solutions that tests alone would miss.
-
November 11, 20265 min read
-
-
-
Featured news
-
CODE@MIT 20262026Many large-scale A/B testing systems leverage normal-normal empirical Bayes models to improve estimates of treatment impacts and launch decisions. We show that in these models maximum likelihood estimates of the variance of true impacts can be sensitive to a small number of extreme observations. We propose a bounded influence maximum likelihood estimator (BIMLE) that downweights such observations in proportion
-
2026Can an LLM read a 30+ page Standard Operating Procedure (SOP) and produce a team of agents that reliably execute it? Industrial SOPs bury workflow logic in prose, leave inter-step dependencies implicit, and scatter referenced entities across sections, forcing planners to parse, sequence, and tool-map within one context window. We argue that the missing primitive is procedural chunking : decomposing a document
-
2026Retrieving relevant pages from visually rich documents, with evidence in charts, tables, and figures rather than body text, remains difficult: a single-pass query against a fixed index misses content in visual elements, and reordering the same candidates cannot recover a page never retrieved. We propose Reflect and Retrieve, an agentic retrieval approach centered on a Reflection Agent. After an initial
-
Concentration beats coverage: Impact-aware expert review for AI-assisted carbon footprint estimationNeurIPS 2026 Workshop on Tackling Climate Change with Machine Learning2026Reducing the rapidly growing carbon footprint of computing at scale will require replacing month-long manual life cycle assessment (LCA) with AI estimates of product carbon footprint (PCF) to inform product design, supplier selection, and portfolio-level decarbonization decisions. These estimates are cheap but often uncertain, while expert review and primary data collection are accurate but costly. This
-
WMT 20262026Large-scale machine-translation (MT) systems are typically evaluated on random samples from a corpus whose distributional composition is an artifact of how it was assembled. Such a sample inherits the phenomena the collection happens to contain rather than the full space a system must handle, spanning rule-governed conventions (terminology, punctuation, currency formatting) and context-dependent phenomena
Collaborations
View allWhether you're a faculty member or student, there are number of ways you can engage with Amazon.
View all