Skip To Main Content
van Deemter on Muck Rack

van Deemter

As seen in: Nature, arXiv, Frontiers, Cancer Research
Is this you? As a journalist, you can create a free Muck Rack account to customize your profile, list your contact preferences, and upload a portfolio of your best work. Claim your profile
Back to van Deemter’s profile

Articles by van Deemter

My Life in Artificial Intelligence: People, anecdotes, and some lessons learnt

|
By van Deemter
| arXiv
arXiv:2504.04142 (cs) View PDF Comments: Subjects: Computation and Language (cs.CL); Artificial Intelligence (cs.AI) Cite as: arXiv:2504.04142 [cs.CL] (or arXiv:2504.04142v1 [cs.CL] for this version) https://doi.org/10.48550/arXiv.2504.04142 Submission history From: Kees van Deemter [ view email] [v1] Sat, 5 Apr 2025 11:26:48 UTC (677 KB) Bibliographic Tools Bibliographic Explorer Toggle Bibliographic Explorer () Connected Papers Toggle Connected Papers (What is Connected Papers?) Litmaps...
Open in Who Shared View Duplicates (202) Issue with article?

[2501.12011] Reference-free Evaluation Metrics for Text Generation: A Survey

|
By van Deemter
| arXiv
arXiv:2501.12011 (cs) View PDF HTML (experimental) Comments: Subjects: Computation and Language (cs.CL) Cite as: arXiv:2501.12011 [cs.CL] (or arXiv:2501.12011v1 [cs.CL] for this version) https://doi.org/10.48550/arXiv.2501.12011 Submission history From: Takumi Ito [ view email] [v1] Tue, 21 Jan 2025 10:05:48 UTC (43 KB) Bibliographic Tools Bibliographic Explorer Toggle Bibliographic Explorer () Connected Papers Toggle Connected Papers (What is Connected Papers?) Litmaps Toggle Litmaps (What...
Open in Who Shared View Duplicates (67) Issue with article?

[2403.04376] Computational Modelling of Plurality and Definiteness in Chinese Noun Phrases

|
By van Deemter, Guanyi Chen, Kees van Deemter
| arXiv
Open in Who Shared Issue with article?

[2402.07432] Intrinsic Task-based Evaluation for Referring Expression Generation

|
By van Deemter
| arXiv
arXivLabs is a framework that allows collaborators to develop and share new arXiv features directly on our website. Both individuals and organizations that work with arXivLabs have embraced and accepted our values of openness, community, excellence, and user data privacy. arXiv is committed to these values and only works with partners that adhere to them. Have an idea for a project that will add value for arXiv's community? Learn more about arXivLabs.
Open in Who Shared View Duplicates (699) Issue with article?

Textual Summarisation of Large Sets: Towards a General Approach

|
By van Deemter
| arXiv
Open in Who Shared Issue with article?

Matrix Stiffness Triggers Lipid Metabolic Cross-talk between Tumor and Stromal Cells to Mediate Bevacizumab Resistance in Colorectal Cancer Liver Metastases

|
By R. J. Zhou, van Deemter, Vander Heiden
| Cancer Research Verified
The average 5-year overall survival (OS) rate for colorectal cancer is 60%, making it the third most prevalent cancer in the world (1). Primary colorectal cancer can be treated surgically; however, patients with metastatic colorectal cancer (mCRC) have a considerably lower chance of surviving.
Open in Who Shared Issue with article?

Models of reference production: How do they withstand the test of time?

|
By van Deemter, Guanyi Chen, Kees van Deemter
| arXiv
In recent years, many NLP studies have focused solely on performance improvement. In this work, we focus on the linguistic and scientific aspects of NLP. We use the task of generating referring expressions in context (REG-in-context) as a case study and start our analysis from GREC, a comprehensive set of shared tasks in English that addressed this topic over a decade ago.
Open in Who Shared Issue with article?

[2305.14020] Does ChatGPT have Theory of Mind?

|
By van Deemter, Kees van Deemter
| arXiv
arXiv:2305.14020 (cs) [Submitted on 23 May 2023 (v1), last revised 13 Sep 2023 (this version, v2)] Download PDF Subjects: Computation and Language (cs.CL); Artificial Intelligence (cs.AI) Cite as: arXiv:2305.14020 [cs.CL] (or arXiv:2305.14020v2 [cs.CL] for this version) https://doi.org/10.48550/arXiv.2305.14020 Submission history From: Bart Holterman [ view email] [v1] Tue, 23 May 2023 12:55:21 UTC (51 KB) [v2] Wed, 13 Sep 2023 11:22:19 UTC (1,783 KB) Bibliographic Tools Bibliographic...
Open in Who Shared View Duplicates (264) Issue with article?

Interpreting Vision and Language Generative Models with Semantic Visual Priors

|
By van Deemter
| Frontiers
Michele Cafagna1* Lina M. Rojas-Barahona2 Kees Van Deemter3 Albert Gatt1, 3 1University of Malta, Malta 2Orange Innovation, France 3Utrecht University, Netherlands The final, formatted version of the article will be published soon. Receive an email when it is updated When applied to Image-to-text models, explainability methods have two challenges. First, they often provide token-by-token explanations namely, they compute a visual explanation for each token of the generated sequence.
Open in Who Shared Issue with article?

Missing Information, Unresponsive Authors, Experimental Flaws: The Impossibility of Assessing the Reproducibility of Previous Human Evaluations in NLP

|
By van Deemter, van der Lee, van Miltenburg, Pablo Mosteiro
| arXiv
arXiv:2305.01633 (cs) Download a PDF of the paper titled Missing Information, Unresponsive Authors, Experimental Flaws: The Impossibility of Assessing the Reproducibility of Previous Human Evaluations in NLP, by Anya Belz and 38 other authors Download PDF Submission history From: Craig Thomson [ view email] [v1] Tue, 2 May 2023 17:46:12 UTC (35 KB) Bibliographic Tools Bibliographic Explorer Toggle Bibliographic Explorer () Litmaps Toggle Litmaps (What is Litmaps?) scite.ai Toggle scite Smart...
Open in Who Shared View Duplicates (517) Issue with article?

[2304.14986] Interpreting Vision and Language Generative Models with Semantic Visual Priors

|
By van Deemter
| arXiv
When applied to Image-to-text models, interpretability methods often provide token-by-token explanations namely, they compute a visual explanation for each token of the generated sequence. Those explanations are expensive to compute and unable to comprehensively explain the model's output. Therefore, these models often require some sort of approximation that eventually leads to misleading explanations.
Open in Who Shared Issue with article?

HL Dataset: Visually-grounded Description of Scenes, Actions and Rationales

|
By van Deemter
| arXiv
[Submitted on 23 Feb 2023 ( v1 ), last revised 25 Sep 2023 (this version, v3)] Title:HL Dataset: Visually-grounded Description of Scenes, Actions and Rationales Download a PDF of the paper titled HL Dataset: Visually-grounded Description of Scenes, Actions and Rationales, by Michele Cafagna and 2 other authors Download PDF Abstract:Current captioning datasets focus on object-centric captions, describing the visible objects in the image, e.g. "people eating food in a park".
Open in Who Shared Issue with article?

The case for open computer programs

|
By De Roeck, Darrel C. Ince, van Deemter, de Bruijn
| Nature Verified
Abstract Scientific communication relies on evidence that cannot be entirely included in publications, but the rise of computational science has added a new layer of inaccessibility. Although it is now accepted that data should be made available on request, the current regulations regarding the availability of software are inconsistent. We argue that, with some exceptions, anything less than the release of source programs is intolerable for results that depend on computation.
Open in Who Shared Issue with article?
Show More