Search Results for author: Carsten Eickhoff

Found 52 papers, 29 papers with code

Self-Supervised Neural Topic Modeling

1 code implementation Findings (EMNLP) 2021 Seyed Ali Bahrainian, Martin Jaggi, Carsten Eickhoff

Topic models are useful tools for analyzing and interpreting the main underlying themes of large corpora of text.

Clustering Topic Models

Controllable Topic-Focused Abstractive Summarization

no code implementations12 Nov 2023 Seyed Ali Bahrainian, Martin Jaggi, Carsten Eickhoff

We show that our model sets a new state of the art on the NEWTS dataset in terms of topic-focused abstractive summarization as well as a topic-prevalence score.

Abstractive Text Summarization

Outlier Dimensions Encode Task-Specific Knowledge

1 code implementation26 Oct 2023 William Rudman, Catherine Chen, Carsten Eickhoff

Representations from large language models (LLMs) are known to be dominated by a small subset of dimensions with exceedingly high variance.

Circuit Component Reuse Across Tasks in Transformer Language Models

no code implementations12 Oct 2023 Jack Merullo, Carsten Eickhoff, Ellie Pavlick

that it is mostly reused to solve a seemingly different task: Colored Objects (Ippolito & Callison-Burch, 2023).

Predictive Uncertainty-based Bias Mitigation in Ranking

1 code implementation18 Sep 2023 Maria Heuss, Daniel Cohen, Masoud Mansoury, Maarten de Rijke, Carsten Eickhoff

Prior work on bias mitigation often assumes that ranking scores, which correspond to the utility that a document holds for a user, can be accurately determined.


One-Versus-Others Attention: Scalable Multimodal Integration

1 code implementation11 Jul 2023 Michal Golovanevsky, Eva Schiller, Akira Nair, Ritambhara Singh, Carsten Eickhoff

Multimodal learning models have become increasingly important as they surpass single-modality approaches on diverse tasks ranging from question-answering to autonomous driving.

Autonomous Driving Question Answering

SSE: A Metric for Evaluating Search System Explainability

1 code implementation16 Jun 2023 Catherine Chen, Carsten Eickhoff

Explainable Information Retrieval (XIR) is a growing research area focused on enhancing transparency and trustworthiness of the complex decision-making processes taking place in modern information retrieval systems.

Decision Making Information Retrieval +1

Stable Anisotropic Regularization

1 code implementation30 May 2023 William Rudman, Carsten Eickhoff

Given the success of Large Language Models (LLMs), there has been considerable interest in studying the properties of model activations.

Enhancing the Ranking Context of Dense Retrieval Methods through Reciprocal Nearest Neighbors

1 code implementation25 May 2023 George Zerveas, Navid Rekabsaz, Carsten Eickhoff

Sparse annotation poses persistent challenges to training dense retrieval models; for example, it distorts the training signal when unlabeled relevant documents are used spuriously as negatives in contrastive learning.

Contrastive Learning Retrieval +1

Neural Summarization of Electronic Health Records

no code implementations24 May 2023 Koyena Pal, Seyed Ali Bahrainian, Laura Mercurio, Carsten Eickhoff

Using nursing notes and discharge summaries from the MIMIC-III dataset, we studied the viability of the automatic generation of various sections of a discharge summary using four state-of-the-art neural network summarization models (BART, T5, Longformer and FLAN-T5).

Language Modelling

CroCoSum: A Benchmark Dataset for Cross-Lingual Code-Switched Summarization

no code implementations7 Mar 2023 Ruochen Zhang, Carsten Eickhoff

Cross-lingual summarization (CLS) has attracted increasing interest in recent years due to the availability of large-scale web-mined datasets and the advancements of multilingual language models.

Unsupervised Multivariate Time-Series Transformers for Seizure Identification on EEG

1 code implementation3 Jan 2023 İlkay Yıldız Potter, George Zerveas, Carsten Eickhoff, Dominique Duncan

Epileptic seizures are commonly monitored through electroencephalogram (EEG) recordings due to their routine and low expense collection.

EEG Electroencephalogram (EEG) +3

Linearly Mapping from Image to Text Space

1 code implementation30 Sep 2022 Jack Merullo, Louis Castricato, Carsten Eickhoff, Ellie Pavlick

Prior work has shown that pretrained LMs can be taught to caption images when a vision model's parameters are optimized to encode images in the language space.

Image Captioning Language Modelling +2

When BERT Fails -- The Limits of EHR Classification

no code implementations26 Jul 2022 Augusto Garcia-Agundez, Carsten Eickhoff

Transformers are powerful text representation learners, useful for all kinds of clinical decision support tasks.

Classification Readmission Prediction

Pretraining on Interactions for Learning Grounded Affordance Representations

1 code implementation *SEM (NAACL) 2022 Jack Merullo, Dylan Ebert, Carsten Eickhoff, Ellie Pavlick

Lexical semantics and cognitive science point to affordances (i. e. the actions that objects support) as critical for understanding and representing nouns and verbs.

Grounded language learning

Multimodal Attention-based Deep Learning for Alzheimer's Disease Diagnosis

1 code implementation17 Jun 2022 Michal Golovanevsky, Carsten Eickhoff, Ritambhara Singh

The objective of this study was to develop a novel multimodal deep learning framework to aid medical professionals in AD diagnosis.

Multi-class Classification Multimodal Deep Learning

Garden-Path Traversal in GPT-2

1 code implementation24 May 2022 William Jurayj, William Rudman, Carsten Eickhoff

In recent years, large-scale transformer decoders such as the GPT-x family of models have become increasingly popular.

Language Modelling

Wasserstein Adversarial Learning based Temporal Knowledge Graph Embedding

no code implementations4 May 2022 Yuanfei Dai, Wenzhong Guo, Carsten Eickhoff

In order to deal with this issue, several temporal knowledge graph embedding (TKGE) approaches have been proposed to integrate temporal and structural information in recent years.

Knowledge Graph Embedding

Image-Like Graph Representations for Improved Molecular Property Prediction

no code implementations20 Nov 2021 Toni Sagayaraj, Carsten Eickhoff

Research into deep learning models for molecular property prediction has primarily focused on the development of better Graph Neural Network (GNN) architectures.

Molecular Property Prediction molecular representation +1

A Novel Corpus of Discourse Structure in Humans and Computers

1 code implementation10 Nov 2021 Babak Hemmatian, Sheridan Feucht, Rachel Avram, Alexander Wey, Muskaan Garg, Kate Spitalnic, Carsten Eickhoff, Ellie Pavlick, Bjorn Sandstede, Steven Sloman

We present a novel corpus of 445 human- and computer-generated documents, comprising about 27, 000 clauses, annotated for semantic clause types and coherence relations that allow for nuanced comparison of artificial and natural discourse modes.

Text Generation

IsoScore: Measuring the Uniformity of Embedding Space Utilization

1 code implementation Findings (ACL) 2022 William Rudman, Nate Gillman, Taylor Rayne, Carsten Eickhoff

We propose IsoScore: a novel tool that quantifies the degree to which a point cloud uniformly utilizes the ambient vector space.

Word Embeddings

The Cross-Lingual Arabic Information REtrieval (CLAIRE) System

no code implementations29 Jul 2021 Zhizhong Chen, Carsten Eickhoff

Despite advances in neural machine translation, cross-lingual retrieval tasks in which queries and documents live in different natural language spaces remain challenging.

Information Retrieval Machine Translation +2

ExpertRank: A Multi-level Coarse-grained Expert-based Listwise Ranking Loss

no code implementations29 Jul 2021 Zhizhong Chen, Carsten Eickhoff

Existing listwise ranking losses treat the candidate document list as a whole unit without further inspection.

Information Retrieval Retrieval

A Modern Perspective on Query Likelihood with Deep Generative Retrieval Models

1 code implementation25 Jun 2021 Oleg Lesota, Navid Rekabsaz, Daniel Cohen, Klaus Antonius Grasserbauer, Carsten Eickhoff, Markus Schedl

In contrast to the matching paradigm, the probabilistic nature of generative rankers readily offers a fine-grained measure of uncertainty.

Passage Re-Ranking Passage Retrieval +3

Not All Relevance Scores are Equal: Efficient Uncertainty and Calibration Modeling for Deep Retrieval Models

1 code implementation10 May 2021 Daniel Cohen, Bhaskar Mitra, Oleg Lesota, Navid Rekabsaz, Carsten Eickhoff

In any ranking system, the retrieval model outputs a single score for a document based on its belief on how relevant it is to a given search query.


TripClick: The Log Files of a Large Health Web Search Engine

1 code implementation14 Mar 2021 Navid Rekabsaz, Oleg Lesota, Markus Schedl, Jon Brassey, Carsten Eickhoff

As such, the collection is one of the few datasets offering the necessary data richness and scale to train neural IR models with a large amount of parameters, and notably the first in the health domain.

Information Retrieval Retrieval

Mining Misdiagnosis Patterns from Biomedical Literature

1 code implementation24 Jun 2020 Cindy Li, Elizabeth Chen, Guergana Savova, Hamish Fraser, Carsten Eickhoff

Diagnostic errors can pose a serious threat to patient safety, leading to serious harm and even death.

Diagnosis Prevalence vs. Efficacy in Machine-learning Based Diagnostic Decision Support

no code implementations24 Jun 2020 Gil Alon, Elizabeth Chen, Guergana Savova, Carsten Eickhoff

Scores fell from 0. 28 for the 50 most prevalent ICD-9-CM codes to 0. 03 for the 1000 most prevalent ICD-9-CM codes.

BIG-bench Machine Learning

Drug-Drug Interaction Prediction with Wasserstein Adversarial Autoencoder-based Knowledge Graph Embeddings

no code implementations15 Apr 2020 Yuanfei Dai, Chenhao Guo, Wenzhong Guo, Carsten Eickhoff

Recently, several knowledge graph embedding approaches have received increasing attention in the DDI domain due to their capability of projecting drugs and interactions into a low-dimensional feature space for predicting links and classifying triplets.

Knowledge Graph Embedding Knowledge Graph Embeddings +1

DC3 -- A Diagnostic Case Challenge Collection for Clinical Decision Support

1 code implementation22 Aug 2019 Carsten Eickhoff, Floran Gmehlin, Anu V. Patel, Jocelyn Boullier, Hamish Fraser

In clinical care, obtaining a correct diagnosis is the first step towards successful treatment and, ultimately, recovery.

On the Effect of Low-Frequency Terms on Neural-IR Models

1 code implementation29 Apr 2019 Sebastian Hofstätter, Navid Rekabsaz, Carsten Eickhoff, Allan Hanbury

Low-frequency terms are a recurring challenge for information retrieval models, especially neural IR frameworks struggle with adequately capturing infrequently observed words.

Passage Retrieval Retrieval +1

Embedding Electronic Health Records for Clinical Information Retrieval

no code implementations13 Nov 2018 Xing Wei, Carsten Eickhoff

Neural network representation learning frameworks have recently shown to be highly effective at a wide range of tasks ranging from radiography interpretation via data-driven diagnostics to clinical decision support.

Information Retrieval Representation Learning +1

Biomedical Question Answering via Weighted Neural Network Passage Retrieval

no code implementations9 Jan 2018 Ferenc Galkó, Carsten Eickhoff

The amount of publicly available biomedical literature has been growing rapidly in recent years, yet question answering systems still struggle to exploit the full potential of this source of data.

Passage Retrieval Question Answering +2

Evaluating Music Recommender Systems for Groups

no code implementations31 Jul 2017 Zsolt Mezei, Carsten Eickhoff

Recommendation to groups of users is a challenging and currently only passingly studied task.

Benchmarking Recommendation Systems

Efficient Parallel Learning of Word2Vec

no code implementations24 Jun 2016 Jeroen B. P. Vuurens, Carsten Eickhoff, Arjen P. de Vries

Since its introduction, Word2Vec and its variants are widely used to learn semantics-preserving representations of words or entities in an embedding space, which can be used to produce state-of-art results for various Natural Language Processing tasks.

Computing Web-scale Topic Models using an Asynchronous Parameter Server

1 code implementation24 May 2016 Rolf Jagerman, Carsten Eickhoff, Maarten de Rijke

Topic models such as Latent Dirichlet Allocation (LDA) have been widely used in information retrieval for tasks ranging from smoothing and feedback methods to tools for exploratory search and discovery.

Information Retrieval Retrieval +1

Probabilistic Bag-Of-Hyperlinks Model for Entity Linking

1 code implementation8 Sep 2015 Octavian-Eugen Ganea, Marina Ganea, Aurelien Lucchi, Carsten Eickhoff, Thomas Hofmann

We demonstrate the accuracy of our approach on a wide range of benchmark datasets, showing that it matches, and in many cases outperforms, existing state-of-the-art methods.

Entity Disambiguation Entity Linking +3

The BladeMistress Corpus: From Talk to Action in Virtual Worlds

no code implementations LREC 2012 Anton Leuski, Carsten Eickhoff, James Ganis, Victor Lavrenko

Thirdly, a joined analysis of both the language and the actions would empower us to build effective modes of the users and their behavior.

Text Classification

Cannot find the paper you are looking for? You can Submit a new open access paper.