Projects

Animal exhibit distinct behavior that dependent their internal states even for the same sensory input. Drosophila larva responds different to odors when starved compared with fed. This internal state dependent oflactory behavior is controled by a topdown feedback to antennal lobe circuit.

Biologically plausible networks models with noisy synaptic update explain the widely observed representational drift.

We extend previous theory to investigate how the interplay of extrinsic noise and intrinsic noise affects early-warning signals near critical transitions.

To achieve efficient coding of odor information in an array of nonlinear olfactory receptors, the odor-receptor sensitivity matrix must be sparse. This sparsity depends on the statistics of environmental odors. We used analytical calclation and extensive numerical simulation to study the optimal sensitivity matrix for recptors with and without spontaneous (background) activity.

Biologically plausible networks models with noisy synaptic update explain the widely observed representational drift.

We built a circuit model of the first olfactory information processing center of fruit fly, which incorporates key features of neuron-neuron interactions such as short-term plasticity and presynaptic inhibition.

Multineuronal imaging of the odor responses in C. elegans and decoding analysis revealed a distinct organization compared with insects and mamammlians.