My main interest focuses on understanding neural processing as a kind of statistical inference. Because most neural systems have to face dynamic environments, this includes the question of how information about time-varying quantities is transmitted in such systems. How much can we know about the input to a group of units by looking at their output behavior? Starting with single units of spiking neurons, this means how much information about their input can be decoded from their output spike trains. We have started to answer these questions by applying methods from information theory and computer simulations to a specific class of mathematical models describing how neurons in the early visual system process the input they receive.
Lochmann, T., Ernst, U.A., and Denève, S.,, Journal of Neuroscience, 32(12), 4179-95 (2012).
Lochmann, T. and Deneve, S., Neural processing as causal inference, Current Opinion in Neurobiology, 21(5), 774-81 (2011).
Lochmann, T. and Deneve, S., Optimal cue combination predict contextual effects on sensory neural responses, Sensory Cue Integration, (2011).
Lochmann, T. and Denève, S., Information transmission with spiking Bayesian neurons, New Journal of Physics, 10, article ID: 055019 (2008).