Preguntar algo…
- User
Preguntar algo…
Assistant
Here are several recent and relevant ERC (European Research Council) grant proposals and funded projects spanning machine learning (ML) and natural language processing (NLP): Example ERC Projects in Machine Learning
Collective Minds (Dr. Sebastian Stich): Funded by an ERC Consolidator Grant, this project is targeted at making machine learning algorithms more sustainable and adaptable. Key goals include training models that more effectively share knowledge, retain new information, and forget outdated data, which is especially advantageous for sensitive domains like healthcare.cispa
Human-AI Coordination (Prof. Jakob Foerster): This ERC Starting Grant supports the development of ML algorithms for human-AI coordination, focusing on interpretable and collaborative agents in scenarios like mixed-autonomy teams or human-robot cooperation.eng.ox
DYNAMICAUS (Dr. Niki Kilbertus): An ERC Starting Grant for causal modeling in complex systems, bridging machine learning, causal inference, and mechanistic modeling to advance the understanding of interventions in scientific and societal contexts.mcml+1
BrainAlign (Dr. Mariya Toneva): A 2025 ERC Starting Grant for developing brain-aligned language models, exploring the intersection of NLP and neuroscience for long-range language understanding.mpi-sws
Example ERC Projects in NLP
DataDivers: This ERC-funded project is aimed at developing frameworks for measuring and utilizing data diversity in NLP datasets. Its primary focus is ensuring fairness and robustness in NLP models by analyzing and leveraging dataset quality rather than just size.cordis.europa
BALM (James Henderson): An ERC Advanced Grant project, BALM is designed to improve the controllability, interpretability, and adaptability of large language models by leveraging Bayesian attention. The outcomes aim to enhance the transparency and reliability of NLP systems for tasks like translation and conversational AI.idiap
PERSONAE: The ERC-funded PERSONAE project is set to revolutionize research on subjective NLP tasks (such as sentiment analysis and abusive language detection) by developing personalizable, privacy-aware NLP models, focusing on tailoring language technologies to individuals.cordis.europa
CulturAL: Targeting global accessibility, the ERC-funded CulturAL project focuses on transferring large language models across different languages, dialects, and cultural contexts to overcome data scarcity in low-resource languages.cordis.europa
INTEGRATOR: This research aims to incorporate demographic factors into natural language processing, resulting in improved algorithms and datasets that account for user diversity.bidsa.unibocconi
Deep Learning for Structured Prediction in NLP: This project bridges deep learning and structured prediction for applications such as machine translation, dependency parsing, and quality estimation, with efforts to incorporate reinforcement learning, latent structure, and weak supervision.cordis.europa
Where to Find ERC Proposals and Summaries ERC project summaries and descriptions—often including published proposal sections—are commonly available through:
The CORDIS Project database (EU’s primary portal for funded research).cordis.europa+3
University and research institute news pages linked to ERC award announcements and press releases.cispa+3
Occasionally, detailed proposals may be uploaded by principal investigators for dissemination and transparency (example: G. Boleda's PDF proposal).gboleda.github
These examples underline current European research frontiers in both ML and NLP, showing a strong emphasis on transparency, fairness, personalization, sustainability, and interdisciplinary integration.