Work Package 2


Networks underlying brain cognition and consciousness

What we do


Our brain is a complex network of neurons able to perform different activities, from a large variety of high-level cognitive functions, to sleep. To improve our understanding of how these processes occur, Work Package 2 will generate data-driven models that carry out cognitive tasks – such as object recognition or decision-making, while expressing realistic brain dynamics in different states (for example, during sleep, awakeness, anaesthesia). This challenging objective requires the coordinated work of a large number of experimental, theoretical and computational research groups, and the results will help to understand not only the healthy brain but also the brain in pathological conditions. The resulting data and models will include a philosophical-ethical framework and be made available to the community.

How we are organised


WP2 is composed of 32 institutions and 40 research groups. WP2 structure is composed of a WP2 leader, WP2 deputy leaders, WP2 managers, WP2 technical coordination and 10 scientific empirical Tasks (T2.1 – T2.6, T2.9, T2.10, T2.12, T2.13), while Task T2.7 is focused on neuroethics and philosophy research, Task T2.8 is focused on management and coordination and Task T2.11 is focused on the EBRAINS integration.



Task T2.1 Data-driven model of multisensory object recognition in cortical systems
Task T2.2 Apical dendritic amplification for cognitive performance: data-driven models and experimental validation
Task T2.3 Global brain dynamics of states leading to cognition: validating models of global brain states and transitions
Task T2.4 From low to high complexity in brain networks
Task T2.5 Multiscale models of brain responsiveness: from single cells to the whole brain
Task T2.6 Mechanisms for conscious and unconscious perception
Task T2.7 Neuroethics and philosophy of cognition and consciousness
Task T2.8 Coordination, management and WP-related outreach
Task T2.9 Influence of Complex Reward Computation and Working Memory Load onto Decision-Making
Task T2.10 Brain Inspired Consciousness
Task T2.11 Technical engineering and scientific integration
Task T2.12 Singularity of the human brain
Task T2.13 Building a prediction model of dopamine drug effects on human cognition



Work Package Leader: Mavi Sanchez-Vives

Work Package Deputy Leader: Cyriel Pennartz

Work Package Managers: Euridice Alvaro and Angelica da Silva Lantyer

Webpage editors: Euridice Alvaro and Angelica da Silva Lantyer


Criteria to analyse broadly relevant issues on the relationships between the emergence of complex networks and consciousness as well as between human and artificial intelligence/cognition.

This report outlines the criteria defined for assessing the theoretical structure of consciousness theories and the relationship between biological and artificial complex networks and consciousness, as well as the feasibility of an artificial emulation of biological intelligence. Specifically, the logical premises for elaborating the criteria for reliable and consistent theories of consciousness are introduced, and then, the theoretical and empirical connections between complexity and consciousness are addressed. Also, the biological and artificial intelligence/cognition are compared.

Human and rodent multiscale datasets integrated in the EBRAINS Knowledge Graph and used to develop representational measures of consciousness.

Two datasets are provided via EBRAINS that allow to study multimodal memory representations as a fundamental aspect impacting consciousness on a multiscale, cross-species level. The first set comprises human ultra-high resolution data and the second rodent electrophysiological data, both covering the medial temporal lobe and cortical areas. Both datasets are acquired with paradigms that require the recall of multimodal information from unimodal cues and are thus suitable to study the nature of multimodal representations in the brain and mechanisms that allow to gain access to them.

Coherent framework of data-driven models that account for stimulus-driven responses in different brain scales. Models of brain signals will be integrated in the LFPy tool.

In this Output, we overview models that were designed at different scales to investigate stimulus-driven responses, from the cellular/circuit level, up to the whole brain. The modelling of signals is also described over these different scales, as well as the LFPy tool. This work spans from SGA1, SGA2 and SGA3 periods of the Human Brain Project.

First release of data-driven mean-field and spiking multi-minicolumn models of state-dependent spontaneous activity and data analysis algorithms for spontaneous activity ranging from wave propagation to measures of network complexity

OP2.9 includes mean-field models, spiking neural networks and analysis pipelines that arise from these models. The Partners involved here are CNRS, ISS, INFN and IDIBAPS.

Concrete empirical, theoretical and behavioural criteria for ascribing consciousness to humans also in clinical contexts (e.g. patients who suffer from disorders of consciousness), animals, and machines suggesting strategies for dealing with key ethical and societal implications of the research results.

The identification of consciousness in other subjects and/or agents, particularly in clinical contexts, is challenging. There is a need for reliable criteria or markers, for instance in the form of indicators that facilitate and justify the attribution of consciousness. The present document summarises relevant work by Task T2.7, focusing on three points in particular: what are the dimensions of consciousness? What are the indicators of consciousness? What are their potential clinical and ethical implications, particularly for the clinical care of patients with Disorders of Consciousness? A note on the question of sentient machines is also included.

Human fMRI, rodent electrophysiology and dendritic calcium imaging datasets will be added to the EBRAINS Knowledge Graph to be used for model validation.

Cognitive operations and consciousness itself depend on the ability of the neocortex to generate internal models of the outside world. We investigate sub-cellular, neuronal and microcircuit mechanisms of cortical feedback processing, towards insights into the biophysics, function and dynamics of feedback inputs, and their role in perception, learning, memory and consciousness. All these topics have been discussed in the document along with the publication of the respective datasets or manuscripts.