Copertina

Subword models struggle with word learning, but surprisal hides it

Fuente: arXiv

Salvato in:

Dettagli Bibliografici
Autori principali:	Bunzeck, Bastian, Zarrieß, Sina
Natura:	Preprint
Pubblicazione:	2025
Soggetti:	Computation and Language
Accesso online:	Acceder al recurso
Tags:	Aggiungi Tag Nessun Tag, puoi essere il primo ad aggiungerne!!

Documenti analoghi

Child-directed speech facilitates production, not comprehension, in BabyLMs
di: Bunzeck, Bastian, et al.
Pubblicazione: (2026)

Do Construction Distributions Shape Formal Language Learning In German BabyLMs?
di: Bunzeck, Bastian, et al.
Pubblicazione: (2025)

Small Language Models Also Work With Small Vocabularies: Probing the Linguistic Abilities of Grapheme- and Phoneme-Based Baby Llamas
di: Bunzeck, Bastian, et al.
Pubblicazione: (2024)

Dialogue Is Not Enough to Make a Communicative BabyLM (But Neither Is Developmentally Inspired Reinforcement Learning)
di: Padovani, Francesca, et al.
Pubblicazione: (2025)

SceneGram: Conceptualizing and Describing Tangrams in Scene Context
di: Junker, Simeon, et al.
Pubblicazione: (2025)

The Frequency Confound in Language-Model Surprisal and Metaphor Novelty
di: Momen, Omar, et al.
Pubblicazione: (2026)

Resilience through Scene Context in Visual Referring Expression Generation
di: Junker, Simeon, et al.
Pubblicazione: (2024)

How Hypocritical Is Your LLM judge? Listener-Speaker Asymmetries in the Pragmatic Competence of Large Language Models
di: Sieker, Judith, et al.
Pubblicazione: (2026)

Rationalizing Transformer Predictions via End-To-End Differentiable Self-Training
di: Brinner, Marc, et al.
Pubblicazione: (2025)

Efficient Scientific Full Text Classification: The Case of EICAT Impact Assessments
di: Brinner, Marc Felix, et al.
Pubblicazione: (2025)

Model Interpretability and Rationale Extraction by Input Mask Optimization
di: Brinner, Marc, et al.
Pubblicazione: (2025)

SemCSE: Semantic Contrastive Sentence Embeddings Using LLM-Generated Summaries For Scientific Abstracts
di: Brinner, Marc, et al.
Pubblicazione: (2025)

LLMs Struggle to Reject False Presuppositions when Misinformation Stakes are High
di: Sieker, Judith, et al.
Pubblicazione: (2025)

SemCSE-Multi: Multifaceted and Decodable Embeddings for Aspect-Specific and Interpretable Scientific Domain Mapping
di: Brinner, Marc, et al.
Pubblicazione: (2025)

Can LLMs Ground when they (Don't) Know: A Study on Direct and Loaded Political Questions
di: Lachenmaier, Clara, et al.
Pubblicazione: (2025)

Talking to a Know-It-All GPT or a Second-Guesser Claude? How Repair reveals unreliable Multi-Turn Behavior in LLMs
di: Lachenmaier, Clara, et al.
Pubblicazione: (2026)

Implicit Causality-biases in humans and LLMs as a tool for benchmarking LLM discourse capabilities
di: Kankowski, Florian, et al.
Pubblicazione: (2025)

Are BabyLMs Deaf to Gricean Maxims? A Pragmatic Evaluation of Sample-efficient Language Models
di: Askari, Raha, et al.
Pubblicazione: (2025)

Reference Games as a Testbed for the Alignment of Model Uncertainty and Clarification Requests
di: Ali, Manar, et al.
Pubblicazione: (2026)

Evaluating Diversity in Automatic Poetry Generation
di: Chen, Yanran, et al.
Pubblicazione: (2024)

Enhancing Domain-Specific Encoder Models with LLM-Generated Data: How to Leverage Ontologies, and How to Do Without Them
di: Brinner, Marc, et al.
Pubblicazione: (2025)

CAIT: A Syntactic Parsing Toolkit for Child-Adult InTeractions
di: Padovani, Francesca, et al.
Pubblicazione: (2026)

Evaluating Subword Tokenization: Alien Subword Composition and OOV Generalization Challenge
di: Batsuren, Khuyagbaatar, et al.
Pubblicazione: (2024)

Distributional Properties of Subword Regularization
di: Cognetta, Marco, et al.
Pubblicazione: (2024)

Lexically Grounded Subword Segmentation
di: Libovický, Jindřich, et al.
Pubblicazione: (2024)

The InviTE Corpus: Annotating Invectives in Tudor English Texts for Computational Modeling
di: Spliethoff, Sophie, et al.
Pubblicazione: (2025)

Are Multimodal Large Language Models Pragmatically Competent Listeners in Simple Reference Resolution Tasks?
di: Junker, Simeon, et al.
Pubblicazione: (2025)

Surprisal and Metaphor Novelty Judgments: Moderate Correlations and Divergent Scaling Effects Revealed by Corpus-Based and Synthetic Datasets
di: Momen, Omar, et al.
Pubblicazione: (2026)

ByteSpan: Information-Driven Subword Tokenisation
di: Goriely, Zébulon, et al.
Pubblicazione: (2025)

Subword Tokenization Strategies for Kurdish Word Embeddings
di: Salehi, Ali, et al.
Pubblicazione: (2025)

Language models struggle with compartmentalization
di: Howe, Thomas Vincent, et al.
Pubblicazione: (2026)

Deep learning models for representing out-of-vocabulary words
di: Lochter, Johannes V., et al.
Pubblicazione: (2020)

The Learning Dynamics of Subword Segmentation for Morphologically Diverse Languages
di: Meyer, Francois, et al.
Pubblicazione: (2025)

Morphological Typology in BPE Subword Productivity and Language Modeling
di: Parra, Iñigo
Pubblicazione: (2024)

On the robustness of modeling grounded word learning through a child's egocentric input
di: Vong, Wai Keen, et al.
Pubblicazione: (2025)

Learning Mutually Informed Representations for Characters and Subwords
di: Wang, Yilin, et al.
Pubblicazione: (2023)

StochasTok: Improving Fine-Grained Subword Understanding in LLMs
di: Sims, Anya, et al.
Pubblicazione: (2025)

Stolen Subwords: Importance of Vocabularies for Machine Translation Model Stealing
di: Zouhar, Vilém
Pubblicazione: (2024)

Tokenization Falling Short: On Subword Robustness in Large Language Models
di: Chai, Yekun, et al.
Pubblicazione: (2024)

The Illusion of Competence: Evaluating the Effect of Explanations on Users' Mental Models of Visual Question Answering Systems
di: Sieker, Judith, et al.
Pubblicazione: (2024)