3 papers
cs.LG2026
Transformers Linearly Represent Highly Structured World Models
Roman Kniazev, Nathanaël Fijalkow
Do transformers, when trained on sequential reasoning traces, build internal models of the underlying task? And if so, does the structure of those internal representations mirror t…
cs.AI2026
Computing the Reachability Value of Posterior-Deterministic POMDPs
Nathanaël Fijalkow, Arka Ghosh, Roman Kniazev +2
Partially observable Markov decision processes (POMDPs) are a fundamental model for sequential decision-making under uncertainty. However, many verification and synthesis problems…
cs.LO2025
A categorical and logical framework for iterated protocols
Eric Goubault, Bernardo Hummes Flores, Roman Kniazev +2
In this article, we show that the now classical protocol complex approach to distributed task solvability of Herlihy et al. can be understood in standard categorical terms. First,…