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
Articles by Christopher Schröder
When Attention Goes Blind: Numerical Failure in ALiBi Positional Encodings
Published on Aug 4 · Submitted by Christopher Schröder on Aug 5 · CORAL NLP Research Abstract We identify a previously overlooked failure mode of ALiBi positional encoding: its linear bias scaling underflows floating-point precision, which zeroes out a large fraction of attention weights and renders the affected attention heads partially blind. We analyze this failure mode, characterize its impact, and examine four mitigation strategies.
Research Portal of the TU Dresden
Reassessing Active Learning Adoption in Contemporary NLP: A Community Survey. Research output: Contribution to book/Conference proceedings/Anthology/Report › Conference contribution › Contributed › peer-review
Have LLMs Made Active Learning Obsolete? Surveying the NLP Community.
Details Originalsprache Englisch Publikationsstatus Veröffentlicht - 2025 No renderer: customAssociatesEventsRenderPortal,dk.atira.pure.api.shared.model.researchoutput.WorkingPaper
No indications of weight gain associated DNA methylation changes in patients with anorexia nervosa
Abstract External IDs PubMedCentral PMC12331989 Keywords Humans, Anorexia Nervosa/genetics, DNA Methylation, Female, Weight Gain/genetics, Adult, CpG Islands, Young Adult, Epigenesis, Genetic, Adolescent, Case-Control Studies
Dominant variants in major spliceosome U4 and U5 small nuclear RNA genes cause neurodevelopmental disorders through splicing disruption - Nature Genetics
|
Abstract The major spliceosome contains five small nuclear RNAs (snRNAs; U1, U2, U4, U5 and U6) essential for splicing. Variants in RNU4-2, encoding U4, cause a neurodevelopmental disorder called ReNU syndrome. We investigated de novo variants in 50 snRNA-encoding genes in a French cohort of 23,649 individuals with rare disorders and gathered additional cases through international collaborations.
By Caroline Nava, Benjamin Cogné, Amandine Santini, Elsa Leitão, François Lecoquierre, Yuyang Chen, Sarah Stenton, Sonja Neuser, Konrad Platzer, Chloé Angelini, Benoît Arveiler, Tahsin Stefan Barakat, Stéphanie Baulac, Claire Beneteau, Stéphane Bézieau, Samantha J. Bryen, Julien Buratti, Almuth Caliebe, John Christodoulou, Andree Delahaye-Duriez, Anne-Sophie Denommé-Pichon, Christèle Dubourg, Salima El Chehadeh, Laurence Faivre, Mélanie Fradin, Svetlana Gorokhova, Mathilde Gras, John M. Greally, Saurav Guha, Tobias B. Haack, Madeleine Harris, Marc-Phillip Hitz, A. Micheil Innes, Celine Jost, Petra Ketteler, Fabian Kilpert, Florian Kraft, Steven S. Laurie, Louis Lebreton, Gaetan Lesca, Jonathan Levy, Agnes Linglart, Jean-Louis Mandel, Víctor Marín, Henri Margot, Johannes A. Mayr, Vincent Michaud, Matthew J. Osmond, Katrin Õunap, Laurent Pasquier, Sandrine Passemard, Melissa Pauly, Christophe Philippe, Annapurna Poduri, Aurora Pujol, Caroline A. Racine, Francis Ramond, André Reis, Rocio Rius, Diana Rodriguez, Clément Sauvestre, Christopher Schröder, Malte Spielmann, Mark Tarnopolsky, Annick Toutain, Frédéric Tran Mau-Them, Julien Van-Gils, Gabriella Vera, André Verdel, Alain Verloes, Marie Vincent, Wojciech P. Galej, Pierre Marijon, Marlene Rio, Séverine Drunat, Nicolas Chatron, Amélie Piton, Gaël Nicolas, Matias Wagner, Rami Abou Jamra, Anne H. O’Donnell-Luria, Nicola Whiffin, Abhilasha Jakhar, Auriane Cospain, Kevin Lüthy, Agatha Schlüter Martin, Harald Surowy, Yoann Vial
|
Nature
Verified
Self-Training for Sample-Efficient Active Learning for Text Classification with Pre-Trained Language Models
Active learning is an iterative labeling process that is used to obtain a small labeled subset, despite the absence of labeled data, thereby enabling to train a model for supervised tasks such as text classification.
Identification and characterisation of pathogenic and non-pathogenic FGF14 repeat expansions - Nature Communications
Abstract Repeat expansions in FGF14 cause autosomal dominant late-onset cerebellar ataxia (SCA27B) with estimated pathogenic thresholds of 250 (incomplete penetrance) and 300 AAG repeats (full penetrance), but the sequence of pathogenic and non-pathogenic expansions remains unexplored. Here, we demonstrate that STRling and ExpansionHunter accurately detect FGF14 expansions from short-read genome data using outlier approaches.
Insane in the vembrane: filtering and transforming VCF/BCF files
Thank you for your interest in spreading the word about bioRxiv. NOTE: Your email address is requested solely to identify you as the sender of this article. Your Email * Your Name * Send To * Enter multiple addresses on separate lines or separate them with commas. Message Subject (Your Name) has forwarded a page to you from bioRxiv Message Body (Your Name) thought you would like to see this page from the bioRxiv website.
Show More
loading
Actions
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.Get in touch with Christopher
Contact Christopher, search articles and posts on X, monitor coverage, and track replies from one place.
Learn more about Muck Rack