Find below EMERGE’s news articles, press releases, digests of scientific publications, as well as other outreach materials.
Publication: I Know How: Combining Prior Policies to Solve New Tasks
28 August 2024
In this work, EMERGE partners from the University of Pisa propose a new framework, I Know How (IKH), which provides a common formalisation to Multi-Task Reinforcement Learning.
Publication: Continuously deep recurrent neural networks
22 August 2024
In this work, EMERGE partners from the University of Pisa and collaborators introduce a new class of recurrent neural models based on a fundamentally different type of topological organisation than the conventionally used deep recurrent networks, and directly inspired by the way cortical networks in the brain process information at multiple temporal scales.
Publication: Temporal Graph ODEs for Irregularly-Sampled Time Series
09 August 2024
In this work, EMERGE partners from the University of Pisa introduce the Temporal Graph Ordinary Differential Equation (TG-ODE) framework, which learns both the temporal and spatial dynamics from graph streams where the intervals between observations are not regularly spaced.
Publication: Investigating over-parameterized randomized graph networks
07 August 2024
In this work, EMERGE partners from the University of Pisa investigate neural models based on graph random features for classification tasks.
Publication: An Experimental Study of Model-based Control for Planar Handed Shearing Auxetics Robots
06 August 2024
In this work, EMERGE partners from the Delft University of Technology and collaborators present a model-based control strategy for planar HSA robots enabling regulation in task space.
Shared awareness could lead to greener, more ethical, and useful smart machines
31 July 2024
The EMERGE project proposes a collaborative shared awareness as a more reliable, energy-efficient, and ethically tractable framework for the coordination between artificial systems and humans than an artificial general intelligence. Read our Press Release!
Publication: Shared Awareness Across Domain-Specific Artificial Intelligence: An Alternative to Domain-General Intelligence and Artificial Consciousness
17 July 2024
In this work, EMERGE partners add their voice to a few ones which propose that investing in specialized AI systems tailored to specific tasks can be overall more effective than developing general intelligence.
Publication: Learning Causal Abstractions of Linear Structural Causal Models
15 July 2024
In this work, EMERGE partners from the University of Pisa tackle issues in modelling causal knowledge for linear causal models with linear abstraction functions.
Publication: Continual pre-training mitigates forgetting in language and vision
01 July 2024
In this work, EMERGE partners from the University of Pisa investigate the characteristics of the Continual Pre-Training scenario, where a model is continually pre-trained on a stream of incoming data and only later fine-tuned to different downstream tasks.
