Skip to content

Publications

72 publications · last updated 2026-05-12 · ORCID · Google Scholar

Authorship convention

In NLP and ML venues, last author typically denotes the senior/PI position; first author is the lead executor (often a student I supervise). In cardiology, food science, and human genetics, the author position has different meanings — typically lead author is the primary investigator and middle authors contribute data, methods, or analysis.

Selected

NLP for Icelandic & Germanic Selected

FoQA: A Faroese question-answering dataset

Simonsen, A., Nielsen, D. S., Einarsson, H. · Proceedings of the third workshop on resources and representations for under-resourced languages and domains (RESOURCEFUL-2025) · 2025

Introduces FoQA, the first dedicated question-answering dataset for Faroese, with extractive and generative variants. The dataset is constructed via translation and adaptation of existing QA benchmarks combined with native-speaker validation, and is accompanied by baseline evaluations of multilingual and Faroese- tuned language models. Establishes a reference benchmark for Faroese reading comprehension. RESOURCEFUL 2025 (NoDaLiDa co-located workshop).

NLP for Icelandic & Germanic Selected

Application of ChatGPT for automated problem reframing across academic domains

Einarsson, H., Lund, S. H., Jónsdóttir, A. H. · Computers and Education: Artificial Intelligence · 2024

Studies the use of ChatGPT for automated problem reframing in academic settings, where the same underlying task is rewritten as a series of alternative problem statements to support student learning and assessment design. Reports on prompting strategies, quality control, and human evaluation across multiple academic domains. First-author paper in Computers and Education: Artificial Intelligence; one of the early systematic studies of generative AI in higher-education problem authoring.

Genetics Selected

Sequence variants associated with BMI affect disease risk through BMI itself

Einarsson, G., Thorleifsson, G., Steinthorsdottir, V., Zink, F., Helgason, H., Olafsdottir, T., Rognvaldsson, S., Tragante, V., Ulfarsson, M. O., Sveinbjornsson, G., others, · Nature Communications · 2024

deCODE-led genome-wide study (Nature Communications 2024) showing that sequence variants associated with body mass index influence downstream disease risk primarily through BMI itself, rather than via independent pleiotropic effects. Uses Mendelian-randomisation and stratified-association analyses across the Icelandic population. Among the few Nature Communications papers in the collection.

NLP for Icelandic & Germanic Selected

Gendered grammar or ingrained bias? exploring gender bias in Icelandic language models

Friðriksdóttir, S. R., Einarsson, H. · Proceedings of the 2024 joint international conference on computational linguistics, language resources and evaluation (LREC-COLING 2024) · 2024

Investigates whether observed gender bias in Icelandic language models reflects grammatical gender or socially ingrained bias. Designs probing tasks that disentangle grammatical agreement from stereotypical role associations, and evaluates several open-source Icelandic LMs (including the IceBERT and GPT-SW3 lines). Finds consistent stereotype-aligned bias beyond what grammar predicts. LREC-COLING 2024.

Clinical AI Selected

Applications of BERT models towards automation of clinical coding in icelandic

Hauksson, H. O., Einarsson, H. · Findings of the association for computational linguistics: NAACL 2024 · 2024

Applies BERT-style models to automate ICD-10 clinical coding in Icelandic hospital records, treating coding as a hierarchical multi-label classification problem. Combines an Icelandic clinical pre-training corpus with multi-task fine-tuning and reports substantial improvements over rule-based and shallow-classifier baselines. The work was awarded the University of Iceland Science Award. Published in Findings of NAACL 2024.

Computer Vision · Natural Sciences Selected

Otolith age determination with a simple computer vision based few-shot learning method

Sigurðardóttir, A. R., Sverrisson, Þ., Jónsdóttir, A., Gudjonsdottir, M., Elvarsson, B. Þ., Einarsson, H. · Ecological Informatics · 2023

Few-shot computer-vision method for fish-age determination from otolith images. The pipeline combines a lightweight feature extractor with a few-shot classifier, achieving accuracy comparable to expert age readings while requiring only a handful of labelled examples per age class. Demonstrated on Atlantic cod otoliths and designed to lower the cost of routine stock-assessment work at fisheries laboratories. Published in Ecological Informatics.

NLP for Icelandic & Germanic Selected

A warm start and a clean crawled corpus-a recipe for good language models

Snæbjarnarson, V., Símonarson, H. B., Ragnarsson, P. O., Ingólfsdóttir, S. L., Jónsson, H., Þorsteinsson, V., Einarsson, H. · Proceedings of the thirteenth language resources and evaluation conference · 2022

Foundational paper for the Icelandic language-model line at the University of Iceland / Miðeind. Demonstrates that warm-starting a Transformer encoder from multilingual weights and continuing pre-training on a carefully cleaned Icelandic web corpus yields strong downstream performance with far less compute than training from scratch. The cleaned crawl and the IceBERT model released alongside the paper became the basis for much of the subsequent Icelandic NLP work in this collection. LREC 2022.

Computational Neuroscience Selected

Bootstrap percolation with inhibition

Einarsson, H., Lengler, J., Mousset, F., Panagiotou, K., Steger, A. · Random Structures & Algorithms · 2019

Studies bootstrap percolation on random graphs in the presence of inhibitory vertices, a generalisation of the classical activation cascade model motivated by cortical excitation-inhibition balance. Introduces a probabilistic framework that captures how the addition of inhibition shifts the activation threshold and the size of the final active set. Published in Random Structures & Algorithms, a premier journal in probabilistic combinatorics; one of the key results of the author's PhD.

All publications

NLP for Icelandic and Germanic languages

NLP for Icelandic & Germanic

Wikiqa-is: Assisted benchmark generation and automated evaluation of icelandic cultural knowledge in llms

Arnardóttir, Þ., Einarsson, E. B., Juto, G. I., Helgason, T. P., Einarsson, H. · Proceedings of the third workshop on resources and representations for under-resourced languages and domains (RESOURCEFUL-2025) · 2025

Introduces WikiQA-IS, an LLM-assisted benchmark for evaluating Icelandic cultural knowledge in language models, with automated evaluation protocols. RESOURCEFUL 2025.

NLP for Icelandic & Germanic

Global piqa: Evaluating physical commonsense reasoning across 100+ languages and cultures

Chang, T. A., Arnett, C., Eldesokey, A., Sadallah, A., Kashar, A., Daud, A., Olanihun, A. G., Mohammed, A. L., Praise, A., Sharma, A. M., others, · arXiv preprint arXiv:2510.24081 · 2025

Co-authored contribution (Icelandic subset) to Global PIQA, a physical-commonsense reasoning benchmark spanning 100+ languages and cultures. arXiv preprint.

NLP for Icelandic & Germanic

MazeEval: a benchmark for testing sequential decision-making in language models

Einarsson, H. · arXiv preprint arXiv:2507.20395 · 2025

MazeEval is a benchmark for testing sequential decision-making in language models through navigation tasks in procedurally generated mazes. Single-author preprint.

NLP for Icelandic & Germanic

Proceedings of the 1st workshop on nordic-baltic responsible evaluation and alignment of language models (NB-REAL 2025)

Einarsson, H., Simonsen, A., Nielsen, D. S. · Proceedings of the 1st workshop on nordic-baltic responsible evaluation and alignment of language models (NB-REAL 2025) · 2025

Edited proceedings of the 1st Workshop on Nordic-Baltic Responsible Evaluation and Alignment of Language Models (NB-REAL 2025), co-located with NoDaLiDa/Baltic-HLT 2025.

NLP for Icelandic & Germanic

Hotter and colder: a new approach to annotating sentiment, emotions, and bias in icelandic blog comments

Friðriksdóttir, S. R., Nielsen, D. S., Einarsson, H. · Proceedings of the joint 25th nordic conference on computational linguistics and 11th baltic conference on human language technologies (NoDaLiDa/baltic-HLT 2025) · 2025

Companion paper for the Hotter and Colder Icelandic sentiment dataset, describing the annotation methodology for sentiment, emotions, and bias in Icelandic blog comments. NoDaLiDa / Baltic-HLT 2025.

NLP for Icelandic & Germanic

Prompt engineering enhances Faroese MT, but only humans can tell

Scalvini, B., Simonsen, A., Debess, I. N., Einarsson, H. · Proceedings of the joint 25th nordic conference on computational linguistics and 11th baltic conference on human language technologies (NoDaLiDa/baltic-HLT 2025) · 2025

Shows that prompt engineering yields measurable improvements in Faroese machine translation, but that automatic metrics fail to capture them - human evaluation is required. NoDaLiDa / Baltic-HLT 2025.

NLP for Icelandic & Germanic

Rethinking low-resource MT: the surprising effectiveness of fine-tuned multilingual models in the LLM age

Scalvini, B., Debess, I. N., Simonsen, A., Einarsson, H. · Proceedings of the joint 25th nordic conference on computational linguistics and 11th baltic conference on human language technologies (NoDaLiDa/baltic-HLT 2025) · 2025

Argues that fine-tuned multilingual NMT models remain surprisingly competitive with large LLMs on low-resource translation, with experiments on Faroese. NoDaLiDa / Baltic-HLT 2025.

NLP for Icelandic & Germanic

Áskoranir og tækifæri með notkun gervigreindar í háskólastarfi: Raddir kennara og hlutverk háskóla. Frumrannsókn á viðhorfi og reynslu kennara við HÍ

Schram, Á. B., Jóhannesdóttir, S., Einarsson, H., Bjarnadóttir, M. K. · Menntakvika 2025: Ráðstefna í menntavísindum · 2025

Icelandic-language companion to the ICERI 2025 paper - a preliminary study of University of Iceland teachers' attitudes and experience with generative AI in higher education. Menntakvika 2025.

NLP for Icelandic & Germanic

Teachers'attitudes and perceptions of the usefulness of ai in academia: How should universities respond to a changing educational landscape?

Schram, Á., Jóhannesdóttir, S., Einarsson, H., Bjarnadóttir, M. K., Rogers, B. L. · ICERI2025 proceedings · 2025

Survey study of teachers' attitudes and perceptions of AI usefulness in academia, asking how universities should respond to a changing educational landscape. ICERI 2025 proceedings.

NLP for Icelandic & Germanic

FoQA: A Faroese question-answering dataset

Simonsen, A., Nielsen, D. S., Einarsson, H. · Proceedings of the third workshop on resources and representations for under-resourced languages and domains (RESOURCEFUL-2025) · 2025

Introduces FoQA, the first dedicated question-answering dataset for Faroese, with extractive and generative variants. The dataset is constructed via translation and adaptation of existing QA benchmarks combined with native-speaker validation, and is accompanied by baseline evaluations of multilingual and Faroese- tuned language models. Establishes a reference benchmark for Faroese reading comprehension. RESOURCEFUL 2025 (NoDaLiDa co-located workshop).

NLP for Icelandic & Germanic

Beyond error categories: A contextual approach of evaluating emerging spell and grammar checkers

Arnardóttir, Þ., Ingólfsdóttir, S. L., Símonarson, H. B., Einarsson, H., Ingason, A. K., Þorsteinsson, V. · Proceedings of the 3rd annual meeting of the special interest group on under-resourced languages@ LREC-COLING 2024 · 2024

Proposes a contextual evaluation methodology for emerging spell- and grammar-checkers that goes beyond fixed error categories, applied to Icelandic systems. SIGUL workshop at LREC-COLING 2024.

NLP for Icelandic & Germanic

Icelandic GPT-SW3 for spell and grammar checking

Arnardóttir, Þ., Ingólfsdóttir, S. L., Ingvarsson Juto, G., Símonarson, H. B., Einarsson, H., Ingason, A. K., Þorsteinsson, V. · Miðeind ehf · 2024

Adapts the GPT-SW3 family of generative language models for Icelandic spell- and grammar-checking, including dataset construction and a fine-tuning protocol. Released by Miðeind ehf.

NLP for Icelandic & Germanic

Good or bad news? Exploring GPT-4 for sentiment analysis for faroese on a public news corpora

Debess, I. N., Simonsen, A., Einarsson, H. · Proceedings of the 2024 joint international conference on computational linguistics, language resources and evaluation (LREC-COLING 2024) · 2024

Evaluates GPT-4 for Faroese sentiment analysis on a public news corpus, comparing prompting strategies and human judgement. LREC-COLING 2024.

NLP for Icelandic & Germanic

Application of ChatGPT for automated problem reframing across academic domains

Einarsson, H., Lund, S. H., Jónsdóttir, A. H. · Computers and Education: Artificial Intelligence · 2024

Studies the use of ChatGPT for automated problem reframing in academic settings, where the same underlying task is rewritten as a series of alternative problem statements to support student learning and assessment design. Reports on prompting strategies, quality control, and human evaluation across multiple academic domains. First-author paper in Computers and Education: Artificial Intelligence; one of the early systematic studies of generative AI in higher-education problem authoring.

NLP for Icelandic & Germanic

Icelandic sentiment corpus (hotter and colder)

Einarsson, H., Friðriksdóttir, S. R., Arnardóttir, Þ. · Háskóli Íslands · 2024

Release of the Icelandic Sentiment Corpus "Hotter and Colder", annotated for nuanced sentiment and adjacent labels in Icelandic blog comments. Distributed via the University of Iceland.

NLP for Icelandic & Germanic

Gendered grammar or ingrained bias? exploring gender bias in Icelandic language models

Friðriksdóttir, S. R., Einarsson, H. · Proceedings of the 2024 joint international conference on computational linguistics, language resources and evaluation (LREC-COLING 2024) · 2024

Investigates whether observed gender bias in Icelandic language models reflects grammatical gender or socially ingrained bias. Designs probing tasks that disentangle grammatical agreement from stereotypical role associations, and evaluates several open-source Icelandic LMs (including the IceBERT and GPT-SW3 lines). Finds consistent stereotype-aligned bias beyond what grammar predicts. LREC-COLING 2024.

NLP for Icelandic & Germanic

Ice and fire: Dataset on sentiment, emotions, toxicity, sarcasm, hate speech, sympathy and more in Icelandic blog comments

Friðriksdóttir, S. R., Simonsen, A., Ásmundsson, A. S., Friðjónsdóttir, G. L., Ingason, A. K., Snæbjarnarson, V., Einarsson, H. · Proceedings of the fourth workshop on threat, aggression & cyberbullying@ LREC-COLING-2024 · 2024

Introduces Ice and Fire, a dataset of Icelandic blog comments annotated for sentiment, emotions, toxicity, sarcasm, hate speech, sympathy, and related labels. Workshop on Threat, Aggression & Cyberbullying at LREC-COLING 2024.

NLP for Icelandic & Germanic

Improving phrase structure parsing for icelandic

Kristjánsdóttir, I. J., Einarsson, H., Ingason, A. K. · 2024

Improves phrase-structure parsing for Icelandic with neural architectures and additional training data. University of Iceland technical report / preprint.

NLP for Icelandic & Germanic

A human perspective on GPT-4 translations: Analysing Faroese to English news and blog text translations

Simonsen, A., Einarsson, H. · Proceedings of the 25th annual conference of the european association for machine translation (volume 1) · 2024

Human-centred analysis of GPT-4 translations between Faroese and English, evaluating news and blog texts with native speakers. EAMT 2024.

NLP for Icelandic & Germanic

Effective entity disambiguation in low-resource languages: a study of icelandic

Eggertsson, V. Á., Jóhannesson, B. G., Einarsson, H., Loftsson, H. · 2023 IEEE/WIC international conference on web intelligence and intelligent agent technology (WI-IAT) · 2023

Examines effective entity disambiguation for low-resource languages through an Icelandic case study, comparing retrieval-based and encoder-based approaches. IEEE/WIC WI-IAT 2023.

NLP for Icelandic & Germanic

The effect of data encoding on relation triplet identification

Friðriksdóttir, S., Einarsson, H. · Proceedings of the 24th nordic conference on computational linguistics (NoDaLiDa) · 2023

Studies how different input encodings affect relation-triplet identification, with experiments comparing flat-text and structure-aware representations. NoDaLiDa 2023.

NLP for Icelandic & Germanic

GameQA: Gamified mobile app platform for building multiple-domain question-answering datasets

Skarphedinsson, N., Gudmundsson, B., Smari, S., Larusdottir, M. K., Einarsson, H., Khan, A., Nyberg, E., Loftsson, H. · Proceedings of the 17th conference of the european chapter of the association for computational linguistics: System demonstrations · 2023

Demos GameQA, a gamified mobile-app platform for collecting multiple-domain question-answering datasets from crowd participants. EACL 2023 System Demonstrations.

NLP for Icelandic & Germanic

Abstractive text summarization for Icelandic

Sverrisson, Þ., Einarsson, H. · Proceedings of the 24th nordic conference on computational linguistics (NoDaLiDa) · 2023

Abstractive text summarization for Icelandic, exploring fine-tuning of multilingual sequence-to-sequence models on a constructed Icelandic news summarization corpus. NoDaLiDa 2023.

NLP for Icelandic & Germanic

Bootstrapping icelandic knowledge graph data

Einarsson, H., Friðriksdóttir, S. R. · ESSLLI · 2022

Proposes bootstrapping techniques for building Icelandic knowledge- graph data from text, presented at the ESSLLI 2022 student session.

NLP for Icelandic & Germanic

Fictionary-based games for language resource creation

Friðriksdóttir, S. R., Einarsson, H. · Proceedings of the 2nd workshop on novel incentives in data collection from people: models, implementations, challenges and results within LREC 2022 · 2022

Proposes fictionary-style games as a low-cost mechanism for crowdsourcing language resources, with experiments on Icelandic. Workshop on Novel Incentives in Data Collection at LREC 2022.

NLP for Icelandic & Germanic

Friðriksdóttir & einarsson (2022), bootstrapping icelandic knowledge graph data

Friðriksdóttir, S. R., Einarsson, H. · University of Amsterdam/Amsterdam University of Applied Sciences · 2022

Master's-style write-up of "Bootstrapping Icelandic Knowledge Graph Data" (University of Amsterdam), a precursor to the ESSLLI proceedings paper of the same title.

NLP for Icelandic & Germanic

MIM-GOLD-EL-entity linking corpus for Icelandic (22.01)

Friðriksdóttir, S. R., Daníelsson, H., Eggertsson, V., Jóhannesson, B. G., Loftsson, H., Einarsson, H. · University of Iceland · 2022

Release of MIM-GOLD-EL (version 22.01), an entity-linking corpus for Icelandic built on top of the MIM-GOLD named-entity collection. University of Iceland.

NLP for Icelandic & Germanic

Reykjavik university question-answering dataset 1 (RUQuAD-1)-version 22.02

Skarphéðinsson, N., Guðmundsson, B., Smári, S., Lárusdóttir, M., Einarsson, H., Kahn, A., Nyberg, E., Loftsson, H. · Reykjavik University · 2022

Release of RUQuAD-1 (version 22.02), the Reykjavik University Question-Answering Dataset for Icelandic - a SQuAD-style reading comprehension dataset collected through the GameQA mobile platform.

NLP for Icelandic & Germanic

Cross-lingual QA as a stepping stone for monolingual open QA in Icelandic

Snæbjarnarson, V., Einarsson, H. · Proceedings of the workshop on multilingual information access (MIA) · 2022

Uses cross-lingual question answering as a stepping stone for monolingual open-domain QA in Icelandic, showing that English retrieval+QA pipelines can be adapted with limited Icelandic supervision. Workshop on Multilingual Information Access (MIA).

NLP for Icelandic & Germanic

Icelandic WinoGrande v1. 0

Snæbjarnarson, V., Símonarson, H. B., Ragnarsson, P. O., Ingólfsdóttir, S. L., Jónsson, H. P., Þorsteinsson, V., Einarsson, H. · Miðeind ehf · 2022

Icelandic WinoGrande v1.0 - an Icelandic adaptation of the WinoGrande commonsense reasoning benchmark, intended for evaluating Icelandic language models. Released by Miðeind ehf.

NLP for Icelandic & Germanic

Natural questions in icelandic

Snæbjarnarson, V., Einarsson, H. · Proceedings of the thirteenth language resources and evaluation conference · 2022

Introduces Natural Questions in Icelandic (NQiI), an open-domain question answering dataset adapted from the English Natural Questions benchmark. Published at LREC 2022.

NLP for Icelandic & Germanic

NQiI v. 1.1

Snæbjarnarson, V., Einarsson, B. T. T., Auðunardóttir, I. I., Sæmundsson, U. I., Bjarnadóttir, H., Gunnarsson, H. V., Einarsson, H. · Háskóli Íslands · 2022

NQiI v1.1 dataset release - an updated version of the Icelandic Natural Questions QA dataset with additional annotations. Distributed via the University of Iceland.

NLP for Icelandic & Germanic

A warm start and a clean crawled corpus-a recipe for good language models

Snæbjarnarson, V., Símonarson, H. B., Ragnarsson, P. O., Ingólfsdóttir, S. L., Jónsson, H., Þorsteinsson, V., Einarsson, H. · Proceedings of the thirteenth language resources and evaluation conference · 2022

Foundational paper for the Icelandic language-model line at the University of Iceland / Miðeind. Demonstrates that warm-starting a Transformer encoder from multilingual weights and continuing pre-training on a carefully cleaned Icelandic web corpus yields strong downstream performance with far less compute than training from scratch. The cleaned crawl and the IceBERT model released alongside the paper became the basis for much of the subsequent Icelandic NLP work in this collection. LREC 2022.

NLP for Icelandic & Germanic

Building an icelandic entity linking corpus

Steinunn Rut Friðriksdóttir, B. G. J., Einarsson, H. · Proceedings of the workshop on dataset creation for lower-resourced languages within the 13th language resources and evaluation conference · 2022

Describes the construction of an Icelandic entity-linking corpus, detailing annotation guidelines and inter-annotator agreement. Dataset Creation for Lower-Resourced Languages workshop at LREC 2022.

NLP for Icelandic & Germanic

Detecting biased language in icelandic: a named entity recognition approach for socially responsible text analysis

Friðriksdóttir, S. R., Einarsson, H. · Workshop on socially responsible language modelling research

Frames detection of biased language in Icelandic as a named-entity recognition task, with annotations targeting socially responsible text analysis. Workshop on Socially Responsible Language Modelling Research.

Computer vision for natural sciences

Computer Vision · Natural Sciences

Temporal aggregation of vision-language features for high-accuracy fish classification in automated monitoring

Silva Martins, J. R., Bárðarson, H., Guðbrandsson, J., Einarsson, H. · Ecological Informatics · 2025

Aggregates temporal vision-language features for high-accuracy fish classification in automated underwater monitoring streams. Ecological Informatics.

Computer Vision · Natural Sciences

Multispectral imaging techniques for evaluating physicochemical-, antioxidant, and sensory properties of Alaria esculenta and Saccharina latissima

Hrólfsdóttir, A. Þ., Sigurðardóttir, A. R., Sveinsdóttir, H. I., Arason, S., Schultz, N., Einarsson, H., Gudjónsdóttir, M. · Algal Research · 2025

Uses multispectral imaging to evaluate physicochemical, antioxidant, and sensory properties of the brown algae Alaria esculenta and Saccharina latissima. Algal Research.

Computer Vision · Natural Sciences

Multispectral imaging as a predictive tool for freshness of whole Atlantic cod: Compared with sensory, chemical and microbiological analysis

Sigurðardóttir, A. R., Sveinsdóttir, H. I., Schultz, N., Einarsson, H., Gudjónsdóttir, M. · Applied Food Research · 2025

Multispectral-imaging method for predicting freshness of whole Atlantic cod, validated against sensory, chemical, and microbiological reference analyses. Applied Food Research.

Computer Vision · Natural Sciences

Sequence segmentation of nematodes in atlantic cod with multispectral imaging data

Sigurðardóttir, A. R., Sveinsdóttir, H. I., Schultz, N., Einarsson, H., Gudjónsdóttir, M. · Foods · 2024

Multispectral-imaging pipeline for sequence segmentation of nematode parasites in Atlantic cod fillets, intended for industrial quality control. Foods (MDPI).

Computer Vision · Natural Sciences

Efficient development of gait classification models for five-gaited horses based on mobile phone sensors

Davíðsson, H. B., Rees, T., Ólafsdóttir, M. R., Einarsson, H. · Animals · 2023

Develops gait-classification models for five-gaited Icelandic horses using mobile-phone inertial sensors, focused on data-efficient training and on-device inference. Published in Animals (MDPI).

Computer Vision · Natural Sciences

Analyzing the effect of the duration of soil warming on subarctic grasslands using high-resolution multispectral images

Hamedpour, A., Sigurdsson, B. D., Peñuelas, J., Filella, I., Einarsson, H., Latré, S., Stefánsson, T. · EGU general assembly conference abstracts · 2023

Conference abstract (EGU General Assembly 2023) analysing the effect of soil-warming duration on subarctic grasslands using high-resolution multispectral imagery from the ForHot experimental site in Iceland.

Computer Vision · Natural Sciences

Otolith age determination with a simple computer vision based few-shot learning method

Sigurðardóttir, A. R., Sverrisson, Þ., Jónsdóttir, A., Gudjonsdottir, M., Elvarsson, B. Þ., Einarsson, H. · Ecological Informatics · 2023

Few-shot computer-vision method for fish-age determination from otolith images. The pipeline combines a lightweight feature extractor with a few-shot classifier, achieving accuracy comparable to expert age readings while requiring only a handful of labelled examples per age class. Demonstrated on Atlantic cod otoliths and designed to lower the cost of routine stock-assessment work at fisheries laboratories. Published in Ecological Informatics.

Computer Vision · Natural Sciences

Drought response analysis on a geothermally warmed Sub-Arctic grassland ecosystem using multispectral drone images

Hamedpour, A., Sigurdsson, B. D., Penuelas, J., Filella, I., Einarsson, H., Latré, S., Stefánsson, T. · 16th international circumpolar remote sensing symposium · 2022

Conference paper analysing drought response in a geothermally warmed subarctic grassland using multispectral drone imagery, from the ForHot site in Iceland. 16th International Circumpolar Remote Sensing Symposium.

Clinical AI and cardiology

Clinical AI

The impact of country income level on the management of heart failure with preserved ejection fraction: an international survey

Ingimarsdóttir, I. J., Saldarriaga, C., Nielsen, N. C., Einarsson, H., Goldfeder, S., Mewton, N., Barasa, A., Basic, C., Oerlemans, M. I., Niederseer, D., others, · ESC Heart Failure · 2026

International survey of how country income level affects HFpEF management practice, with the Icelandic Heart Failure Registry as one of the national data sources. ESC Heart Failure.

Clinical AI

Enhancing heart failure care through registries: lessons learned from Nordic Countries’ implementation of guidelines

Myhre, P. L., Lindberg, F., Løgstrup, B. B., Ingimarsdottir, I. J., Vistnes, M., Schjødt, I., Ljungman, C., Savarese, G., Dridi, N. P., Einarsson, H., others, · European Journal of Heart Failure · 2026

Nordic-cohort study of how national heart-failure registries can enhance guideline-directed care, with contributions from the Icelandic Heart Failure Registry. European Journal of Heart Failure.

Clinical AI

Unravelling ATTR cardiac amyloidosis in iceland: a nationwide epidemiological study

Bergmann, H. M., Úlfarsson, Æ. Ö., Einarsson, H., Jóhannsdóttir, H. X., Gunnarsson, G. T., Hrafnkelsdóttir, T. J., Andersen, K., Ingimarsdóttir, I. J. · Cardiology · 2025

Nationwide epidemiological study of ATTR cardiac amyloidosis in Iceland, leveraging the Icelandic Heart Failure Registry to estimate prevalence and characterise the patient population. Cardiology journal (Karger).

Clinical AI

Survival impact of amiodarone therapy in heart failure: insights from the Icelandic Heart Failure registry

Gudjonsson, P., Andersen, K., Einarsson, H., Ingimarsdottir, I. · European Heart Journal · 2025

Conference abstract (European Heart Journal supplement) examining the survival impact of amiodarone therapy in heart-failure patients using the Icelandic Heart Failure Registry.

Clinical AI

Differences in heart failure with preserved ejection fraction management between care providers: an international survey

Guidetti, F., Giraldo, C. I. S., Shchendrygina, A., Kida, K., Niederseer, D., Basic, C., Rainer, P. P., Załęska-Kocięcka, M., Ogola, E., Mohty, D., others, · European Journal of Heart Failure · 2025

Co-authored international survey on differences in HFpEF management between care providers, contributing the Icelandic Heart Failure Registry data. European Journal of Heart Failure.

Clinical AI

Factors influencing medication adherence in heart failure patients—a survey among cardiac healthcare providers

Gunnthorsdottir, I., Almarsdottir, A. B., Andersen, K., Gunnarsdottir, A. I., Svansdottir, E., Einarsson, H., Ingimarsdottir, I. J. · Clinical Pharmacology & Therapeutics · 2025

Survey study of factors that cardiac healthcare providers believe influence medication adherence in heart-failure patients. Clinical Pharmacology & Therapeutics.

Clinical AI

Diagnostic and therapeutic practice for HFpEF across continents and regions: An international survey

Ingimarsdottir, I. J., Vishram-Nielsen, J. K., Einarsson, H., Goldfeder, S., Mewton, N., Barasa, A., Basic, C., Oerlemans, M. I., Niederseer, D., Shchendrygina, A., others, · ESC Heart Failure · 2025

Journal version of the international HFpEF practice survey, with the Icelandic Heart Failure Registry providing one of the national data contributions. ESC Heart Failure.

Clinical AI

Factors believed by cardiac care professionals to influence medication adherence in heart failure patients

Gunnthorsdottir, I., Andersen, K., Almarsdottir, A., Gunnarsdottir, A., Svansdottir, E., Einarsson, H., Ingimarsdottir, I. · European journal of heart failure · 2024

Conference abstract (EJHF) on factors believed by cardiac care professionals to influence medication adherence in heart failure patients.

Clinical AI

Real-life management of heart failure: evaluation of medication optimisation in different phenotypes

Hauksdottir, H., Einarsson, H., Gunnthorsdottir, I., Gunnarsson, P., Ingimarsdottir, I. · European journal of heart failure · 2024

Conference abstract (EJHF) evaluating real-life medication optimisation in heart-failure patients across different phenotypes, using the Icelandic Heart Failure Registry.

Clinical AI

Applications of BERT models towards automation of clinical coding in icelandic

Hauksson, H. O., Einarsson, H. · Findings of the association for computational linguistics: NAACL 2024 · 2024

Applies BERT-style models to automate ICD-10 clinical coding in Icelandic hospital records, treating coding as a hierarchical multi-label classification problem. Combines an Icelandic clinical pre-training corpus with multi-task fine-tuning and reports substantial improvements over rule-based and shallow-classifier baselines. The work was awarded the University of Iceland Science Award. Published in Findings of NAACL 2024.

Clinical AI

Diagnostic and therapeutic practice for heart failure with preserved ejection fraction around the world: an international survey

Ingimarsdottir, I., Vishram-Nielsen, J., Einarsson, H., Goldfeder, S., Mewton, N., Barasa, A., Basic, C., Oerlemans, M., Niederseer, D., Gustafsson, F., others, · European journal of heart failure · 2024

Conference abstract (EJHF) of the international HFpEF practice survey contributing Icelandic registry data.

Clinical AI

The Icelandic Heart Failure Registry—A nationwide assessment tool for HF care and intervention in HF treatment

Ingimarsdóttir, I. J., Hansen, J. S., Bergmann, H. M., Einarsson, H. · ESC Heart Failure · 2024

Describes the Icelandic Heart Failure Registry as a nationwide tool for assessing heart-failure care and tracking treatment interventions. ESC Heart Failure.

Clinical AI

Diagnostic and therapeutic practice for Heart Failure with preserved ejection fraction around the world: An international survey

Saldarriaga, C., Gracia, S. S. G., Mejia, M. I. P., Shchendrygina, A., Kida, K., Fauvel, C., Zaleska-Kociecka, M., Mapelli, M., Einarsson, H., Guidetti, F., others, · Current problems in cardiology · 2024

Co-authored international survey of diagnostic and therapeutic practice for heart failure with preserved ejection fraction (HFpEF), contributing the Icelandic Heart Failure Registry data. Current Problems in Cardiology / EJHF (a related conference abstract shares the same citation key).

Clinical AI

Diagnostic and therapeutic practices for Heart Failure with preserved ejection fraction around the world: an international survey

Saldarriaga, C., Ines, C., Goldfeder, S., Palacio, M., Shchendrygina, A., Guidetti, F., Vishram-Nielsen, J., Ingimarsdottir, I., Einarsson, H., Kida, K., others, · European journal of heart failure · 2024

Co-authored international survey of diagnostic and therapeutic practice for heart failure with preserved ejection fraction (HFpEF), contributing the Icelandic Heart Failure Registry data. Current Problems in Cardiology / EJHF (a related conference abstract shares the same citation key).

Clinical AI

Epidemiology of acute heart failure admissions in Iceland: phenotype distribution and prognostic implications

Agnarsson, H., Hansen, J., Gunnarsson, D., Einarsson, H., Gunnthorsdottir, I., Ingimarsdottir, I. · European heart journal · 2023

Conference abstract (European Heart Journal supplement) on the epidemiology of acute heart failure admissions in Iceland, examining phenotype distribution and prognostic implications using national registry data.

Clinical AI

Acute heart failure: causes and outcomes by stages of chronic kidney disease and heartfailure phenotypes

Gunnarsson, D. Á., Ingimarsdóttir, I. J., Einarsson, H., Snædal, S. · ESC heart failure 2023 · 2023

Conference abstract (ESC Heart Failure 2023) on causes and outcomes of acute heart failure stratified by chronic kidney disease stage and heart-failure phenotype, using the Icelandic Heart Failure Registry.

Human genetics and computational biology

Genetics

Variant in a taste receptor locus tied to changes in the use of insomnia medication

Einarsson, G., Arnason, H. K., Gisladottir, R. S., Bjornsdottir, G., Thorgeirsson, T. E., Skuladottir, A., Walters, G. B., Saevarsdottir, S., Magnusson, M. K., Halldorsson, G. H., others, · Biological Psychiatry Global Open Science · 2025

deCODE-led study linking a variant at a taste-receptor locus to changes in the use of insomnia medication. Biological Psychiatry Global Open Science.

Genetics

Sequence variants associated with BMI affect disease risk through BMI itself

Einarsson, G., Thorleifsson, G., Steinthorsdottir, V., Zink, F., Helgason, H., Olafsdottir, T., Rognvaldsson, S., Tragante, V., Ulfarsson, M. O., Sveinbjornsson, G., others, · Nature Communications · 2024

deCODE-led genome-wide study (Nature Communications 2024) showing that sequence variants associated with body mass index influence downstream disease risk primarily through BMI itself, rather than via independent pleiotropic effects. Uses Mendelian-randomisation and stratified-association analyses across the Icelandic population. Among the few Nature Communications papers in the collection.

Genetics

Variants at the interleukin 1 gene locus and pericarditis

Thorolfsdottir, R. B., Jonsdottir, A. B., Sveinbjornsson, G., Aegisdottir, H. M., Oddsson, A., Stefansson, O. A., Halldorsson, G. H., Saevarsdottir, S., Thorleifsson, G., Stefansdottir, L., others, · JAMA cardiology · 2024

deCODE-style genome-wide association study linking variants at the interleukin 1 gene locus to pericarditis risk. JAMA Cardiology 2024.

Computational neuroscience

Computational Neuroscience

Bootstrap percolation with inhibition

Einarsson, H., Lengler, J., Mousset, F., Panagiotou, K., Steger, A. · Random Structures & Algorithms · 2019

Studies bootstrap percolation on random graphs in the presence of inhibitory vertices, a generalisation of the classical activation cascade model motivated by cortical excitation-inhibition balance. Introduces a probabilistic framework that captures how the addition of inhibition shifts the activation threshold and the size of the final active set. Published in Random Structures & Algorithms, a premier journal in probabilistic combinatorics; one of the key results of the author's PhD.

Computational Neuroscience

The linear hidden subset problem for the (1+ 1) EA with scheduled and adaptive mutation rates

Einarsson, H., Lengler, J., Gauy, M. M., Meier, F., Mujika, A., Steger, A., Weissenberger, F. · Proceedings of the genetic and evolutionary computation conference · 2018

Analyses the (1+1) evolutionary algorithm on the linear hidden subset problem with scheduled and adaptive mutation rates. GECCO 2018 proceedings.

Computational Neuroscience

A hippocampal model for behavioral time acquisition and fast bidirectional replay of spatio-temporal memory sequences

Matheus Gauy, M., Lengler, J., Einarsson, H., Meier, F., Weissenberger, F., Yanik, M. F., Steger, A. · Frontiers in neuroscience · 2018

Presents a hippocampal model for behavioral time acquisition and fast bidirectional replay of spatio-temporal memory sequences. Frontiers in Neuroscience.

Computational Neuroscience

On the origin of lognormal network synchrony in CA1

Weissenberger, F., Einarsson, H., Matheus Gauy, M., Meier, F., Mujika, A., Lengler, J., Steger, A. · Hippocampus · 2018

Investigates the origin of lognormal network synchrony in CA1, relating heavy-tailed firing distributions to recurrent network structure. Hippocampus journal.

Computational Neuroscience

Fast emergence of neural structures

Einarsson, H. · ETH Zurich · 2017

PhD thesis (ETH Zurich) on the fast emergence of neural structures. Develops mathematical models of how cortical assemblies, synfire chains, and other functional structures can self-organise rapidly from spiking dynamics and local plasticity rules.

Computational Neuroscience

A model of fast hebbian spike latency normalization

Einarsson, H., Gauy, M. M., Lengler, J., Steger, A. · Frontiers in computational neuroscience · 2017

Proposes a model of fast Hebbian spike-latency normalization, in which timing-dependent plasticity rapidly equalises spike latencies across a population. Frontiers in Computational Neuroscience.

Computational Neuroscience

Long synfire chains emerge by spike-timing dependent plasticity modulated by population activity

Weissenberger, F., Meier, F., Lengler, J., Einarsson, H., Steger, A. · International journal of neural systems · 2017

Shows that long synfire chains can emerge from spike-timing dependent plasticity modulated by population activity, providing a mechanistic account of stable sequence representations. International Journal of Neural Systems.

Computational Neuroscience

Connectivity thresholds for bounded size rules

Lengler, J., Mousset, F., Panagiotou, K., Steger, A., Einarsson, H. · Annals of Applied Probability · 2016

Establishes connectivity thresholds for bounded-size rules in random graph processes, a contribution to the theory of Achlioptas-style edge-selection processes. Annals of Applied Probability.

Computational Neuroscience

A high-capacity model for one shot association learning in the brain

Einarsson, H., Lengler, J., Steger, A. · Frontiers in computational neuroscience · 2014

A high-capacity model for one-shot associative learning in cortical networks, showing how sparse-coding regimes extend the classical Hopfield capacity bound. Published in Frontiers in Computational Neuroscience.

Computational Neuroscience

Hippocam

Porter, B. S., Schmidt, R., Bilkey, D. K., Mallei, A., Ieraci, A., Corna, S., Tardito, D., Lee, F. S., Popoli, M., Buckley, M. G., others,

Conference contribution related to hippocampal modelling (collaboration with the Hippocam consortium).