arXiv subjects

Anna Kozlova

Publications and source records attributed to Anna Kozlova.

4 recordsLinked to original sources

arXiv2026

Performance of Clinical AI System and Physicians and Frontier Language Models in primary care diagnostics

Clinical AI evaluation should encompass diagnosis and management after adaptive information gathering. We compared Doctorina, eight physicians and four standalone frontier language models in 150 synthetic Polish-language primary-care consultations. Doctorina achieved 82.0% Top-1 concordance versus 57.0% for physicians (difference, 25.0 percentage points; 95% confidence interval, 17.7-32.7) and 97.3% versus 85.0% primary-or-reference-differential concordance. Across 149 case pairs, normalized workup and treatment scores were 89.4 versus 66.9 and 83.7 versus 61.2. Doctorina had the highest diagnostic point estimates among all six groups; Kimi K3 ranked next, while Claude Opus 5 led the closely spaced management estimates of Opus, Doctorina and Kimi. A second Doctorina execution reproduced the advantages over physicians across all outcomes. Doctorina's advantage over physicians therefore extended from primary-diagnosis selection to higher-rated diagnostic workup and initial treatment after adaptive consultation.

cs.CL↗

arXiv2026

Doctorina MedBench: A Dialogue-Based Benchmark and Evaluation Framework for Agent-Based Medical AI

We present Doctorina MedBench, an evaluation framework for agent-based medical AI based on the simulation of physician-patient interactions. Unlike traditional medical benchmarks that rely on solving standardized test questions, the proposed approach models a multi-step clinical dialogue in which an AI system must collect medical history, analyze available synthetic attachments when present, formulate differential diagnoses, and provide diagnostic and management recommendations. System performance is evaluated across separate task-level domains, including diagnosis, differential diagnosis, treatment, safety-critical condition handling, and dialogue-step behavior; the broader D.O.T.S. framework is used as a supplementary summary for diagnosis, observations/investigations, treatment, and step count. The framework also supports testing and quality-monitoring workflows intended to identify changes in model behavior during development. It supports safety-oriented cases, category-based sampling of synthetic clinical scenarios, and regression-style comparisons across system versions. In the reported study, the analyzed paired complete-case cohort consisted of 254 physician-authored synthetic clinical cases retained from 261 attempted case identifiers. The evaluation metrics are intended for comparative research on interactive medical AI systems and for studying clinical reasoning workflows in synthetic dialogue settings. Our results suggest that simulated clinical dialogue can provide a complementary assessment setting to traditional examination-style benchmarks, while the reported findings do not establish independent clinical validity, clinical effectiveness, or readiness for real-world deployment.

cs.CL↗

arXiv2020

Hamiltonian decomposition and verifying vertex adjacency in 1-skeleton of the traveling salesperson polytope by variable neighborhood search

We consider a Hamiltonian decomposition problem of partitioning a regular graph into edge-disjoint Hamiltonian cycles. A sufficient condition for vertex adjacency in the 1-skeleton of the traveling salesperson polytope can be formulated as the Hamiltonian decomposition problem in a 4-regular multigraph. We introduce a heuristic general variable neighborhood search algorithm for this problem based on finding a vertex-disjoint cycle cover of the multigraph through reduction to perfect matching and several cycle merging operations. The algorithm has a one-sided error: the answer "not adjacent" is always correct, and was tested on random directed and undirected Hamiltonian cycles and on pyramidal tours.

math.CO↗

arXiv2019

Simulated annealing approach to verify vertex adjacencies in the traveling salesperson polytope

We consider 1-skeletons of the symmetric and asymmetric traveling salesperson polytopes whose vertices are all possible Hamiltonian tours in the complete directed or undirected graph, and the edges are geometric edges or one-dimensional faces of the polytope. It is known that the question whether two vertices of the symmetric or asymmetric traveling salesperson polytopes are nonadjacent is NP-complete. A sufficient condition for nonadjacency can be formulated as a combinatorial problem: if from the edges of two Hamiltonian tours we can construct two complementary Hamiltonian tours, then the corresponding vertices of the traveling salesperson polytope are not adjacent. We consider a heuristic simulated annealing approach to solve this problem. It is based on finding a vertex-disjoint cycle cover and a perfect matching. The algorithm has a one-sided error: the answer "not adjacent" is always correct, and was tested on random and pyramidal Hamiltonian tours.

math.CO↗