How do neurons dynamically encode and treat information? Each neuron communicates with its distinctive language made of long silences intermitted by occasional spikes. The spikes are prompted by the pooled effect of a population of pre-synaptic neurons. To understand the operation made by single neurons is to create a quantitative description of their dynamics. The results presented in this thesis describe the necessary elements for a quantitative description of single neurons. Almost all chapters can be unified under the theme of adaptation. Neuronal adaptation plays an important role in the transduction of a given stimulation into a spike train. The work described here shows how adaptation is brought by every spike in a stereotypical fashion. The spike-triggered adaptation is then measured in three main types of cortical neurons. I analyze in detail how the different adaptation profiles can reproduce the diversity of firing patterns observed in real neurons. I also summarize the most recent results concerning the spike-time prediction in real neurons, resulting in a well-founded single-neuron model. This model is then analyzed to understand how populations can encode time-dependent signals and how time-dependent signals can be decoded from the activity of populations. Finally, two lines of investigation in progress are described, the first expands the study of spike-triggered adaptation on longer time scales and the second extends the quantitative neuron models to models with active dendrites.