2 citations · 2 across the 5 of their papers we have counts for
5 papers
A cubical formalisation of topos causal models: intervention, forcing, and a contextuality obstruction
Karen Sargsyan
Topos causal models recast causal inference inside a topos: a causal world is a presheaf, an intervention is a sub-model named by a characteristic map into the subobject classifier…
A cubical formalisation of conditional independence, Bayesian conditioning, and Pearl's d-separation soundness
Karen Sargsyan
The standard convex-algebra interchange axiom, common to probability-monad formalisations since Stone, is provably too weak to support full Bayesian conditioning. We make this prec…
Functorial Neural Architectures from Higher Inductive Types
Karen Sargsyan
Neural networks often learn the parts of a task but fail on novel combinations of those parts. We argue that this failure is architectural: a decoder generalizes compositionally on…
Localizing Preference Aggregation Conflicts: A Graph-Theoretic Approach Using Sheaves
Karen Sargsyan
We introduce a graph-theoretic framework based on discrete sheaves to diagnose and localize inconsistencies in preference aggregation and, more broadly, in the fusion of partial ra…
Structural Enforcement of Statistical Rigor in AI-Driven Discovery: A Functional Architecture
Karen Sargsyan
AI-Scientist systems risk manufacturing spurious discoveries through uncontrolled multiple testing. We present a functional architecture that enforces statistical rigor at two leve…