Abstract:The functional organization of the perisylvian language network was examined using a functional MRI (fMRI) adaptation paradigm with spoken sentences. In Experiment 1, a given sentence was presented every 14.4 s and repeated two, three, or four times in a row. The study of the temporal properties of the BOLD response revealed a temporal gradient along the dorsal-ventral and rostral-caudal directions: From Heschl's gyrus, where the fastest responses were recorded, responses became increasingly slower toward the posterior part of the superior temporal gyrus and toward the temporal poles and the left inferior frontal gyrus, where the slowest responses were observed. Repetition induced a decrease in amplitude and a speeding up of the BOLD response in the superior temporal sulcus (STS), while the most superior temporal regions were not affected. In Experiment 2, small blocks of six sentences were presented in which either the speaker voice or the linguistic content of the sentence, or both, were repeated. Data analyses revealed a clear asymmetry: While two clusters in the left superior temporal sulcus showed identical repetition suppression whether the sentences were produced by the same speaker or different speakers, the homologous right regions were sensitive to sentence repetition only when the speaker voice remained constant. Thus, hemispheric left regions encode linguistic content while homologous right regions encode more details about extralinguistic features like speaker voice. The results demonstrate the feasibility of using sentence-level adaptation to probe the functional organization of cortical language areas.

Summary. Expectation propagation (EP) is a widely successful algorithm for variational inference. EP is an iterative algorithm used to approximate complicated distributions, typically to find a Gaussian approximation of posterior distributions. In many applications of this type, EP performs extremely well. Surprisingly, despite its widespread use, there are very few theoretical guarantees on Gaussian EP, and it is quite poorly understood. To analyse EP, we first introduce a variant of EP: averaged EP, which operates on a smaller parameter space. We then consider averaged EP and EP in the limit of infinite data, where the overall contribution of each likelihood term is small and where posteriors are almost Gaussian. In this limit, we prove that the iterations of both averaged EP and EP are simple: they behave like iterations of Newton's algorithm for finding the mode of a function. We use this limit behaviour to prove that EP is asymptotically exact, and to obtain other insights into the dynamic behaviour of EP, e.g. that it may diverge under poor initialization exactly like Newton's method. EP is a simple algorithm to state, but a difficult one to study. Our results should facilitate further research into the theoretical properties of this important method.

Skilled behavior often displays signatures of Bayesian inference. In order for the brain to implement the required computations, neuronal activity must carry accurate information about the uncertainty of sensory inputs. Two major approaches have been proposed to study neuronal representations of uncertainty. The first one, the Bayesian decoding approach, aims primarily at decoding the posterior probability distribution of the stimulus from population activity using Bayes’ rule, and indirectly yields uncertainty estimates as a by-product. The second one, which we call the correlational approach, searches for specific features of neuronal activity (such as tuning-curve width and maximum firing-rate) which correlate with uncertainty. To compare these two approaches, we derived a new normative model of sound source localization by Interaural Time Difference (ITD), that reproduces a wealth of behavioral and neural observations. We found that several features of neuronal activity correlated with uncertainty on average, but none provided an accurate estimate of uncertainty on a trial-by-trial basis, indicating that the correlational approach may not reliably identify which aspects of neuronal responses represent uncertainty. In contrast, the Bayesian decoding approach reveals that the activity pattern of the entire population was required to reconstruct the trial-to-trial posterior distribution with Bayes’ rule. These results suggest that uncertainty is unlikely to be represented in a single feature of neuronal activity, and highlight the importance of using a Bayesian decoding approach when exploring the neural basis of uncertainty.

No abstract