Papers for

stochastic control engineers

Papers whose findings have a practical use for this group, as judged from the abstract. Open a paper to read what it means in practice.

Risk sensitive schrödinger bridge does not fit kl projection framework

The Risk-Sensitive Schrödinger Bridge: Is Not a KL Projection

Abstract: The Schrödinger bridge owes its computational power to a single structural fact: by Girsanov's theorem the controlled problem is a Kullback--Leibler (KL) projection onto a fixed reference measure, solvable by alternating projections. This letter shows that the fact does not survive risk sensitivity. When the expected path cost is replaced by the entropic risk measure and both endpoint marginals are kept as hard constraints, the resulting fixed-point bridge value $J_θ$ (the soft-problem value at the multiplier that enforces the terminal constraint) admits no representation as a constrained KL minimum against any fixed path-space reference with a regular endpoint law (a class strictly larger than the uniformly elliptic diffusion references: no Markov property is required), even allowing an additive normalisation depending on the initial marginal. Moreover, no single reference generates the one-parameter family in the risk parameter. The obstruction is computed in closed form: the Gaussian bridge value violates, by exactly $θ/2$, a heat equation that any Gaussian smoothing of a fixed endpoint density must obey. In place of the projection, the theory rests on a terminal-multiplier fixed point and an asymmetric factorisation penalising the score energy of the backward factor.

Wed 23 SeptInformation TheoryArtificial Intelligence
The gist
The Schrödinger bridge problem usually relies on a mathematical trick called KL projection to find solutions, which makes it easier to compute. The authors show that when you add risk sensitivity—a way to factor in uncertainty and worst-case scenarios—this trick no longer works. They prove that the new problem cannot be rewritten as a KL projection and explain exactly why using Gaussian examples. Instead, the problem requires different mathematical tools to solve.
Open → 2609.27250v1