Cover Image

The Strawberry Problem: Emergence of Character-level Understanding in Tokenized Language Models

Fuente: arXiv

Saved in:

Bibliographic Details
Main Authors:	Cosma, Adrian, Ruseti, Stefan, Radoi, Emilian, Dascalu, Mihai
Format:	Preprint
Published:	2025
Subjects:	Computation and Language
Online Access:	Acceder al recurso
Tags:	Add Tag No Tags, Be the first to tag this record!

Similar Items

Training Language Models with homotokens Leads to Delayed Overfitting
by: Cosma, Adrian, et al.
Published: (2026)

How Hard is this Test Set? NLI Characterization by Exploiting Training Dynamics
by: Cosma, Adrian, et al.
Published: (2024)

Value-Aware Numerical Representations for Transformer Language Models
by: Dutulescu, Andreea, et al.
Published: (2026)

A Retrieval-Based Approach to Medical Procedure Matching in Romanian
by: Niculae, Andrei, et al.
Published: (2025)

What Makes a Good Doctor Response? A Study on Text-Based Telemedicine
by: Cosma, Adrian, et al.
Published: (2026)

Neural Grammatical Error Correction for Romanian
by: Cotet, Teodor-Mihai, et al.
Published: (2026)

RoMath: A Mathematical Reasoning Benchmark in Romanian
by: Cosma, Adrian, et al.
Published: (2024)

Dr.Copilot: A Multi-Agent Prompt Optimized Assistant for Improving Patient-Doctor Communication in Romanian
by: Niculae, Andrei, et al.
Published: (2025)

On Model and Data Scaling for Skeleton-based Self-Supervised Gait Recognition
by: Cosma, Adrian, et al.
Published: (2025)

The Paradox of Motion: Evidence for Spurious Correlations in Skeleton-based Gait Recognition Models
by: Cătrună, Andy, et al.
Published: (2024)

Spatial Colour Mixing Illusions as a Perception Stress Test for Vision-Language Models
by: Basoc, Nicoleta-Nina, et al.
Published: (2026)

MoME: Estimating Psychological Traits from Gait with Multi-Stage Mixture of Movement Experts
by: Cǎtrunǎ, Andy, et al.
Published: (2025)

CrossGaze: A Strong Method for 3D Gaze Estimation in the Wild
by: Cătrună, Andy, et al.
Published: (2024)

GaitPT: Skeletons Are All You Need For Gait Recognition
by: Catruna, Andy, et al.
Published: (2023)

Database-Agnostic Gait Enrollment using SetTransformers
by: Basoc, Nicoleta, et al.
Published: (2025)

"Înţelegi Româneşte?'' A Recipe for Romanian Vision-Language Models
by: Masala, Mihai, et al.
Published: (2026)

An In-Vitro Study on Cross-Lingual Generalization in Language Models
by: Cosma, Adrian
Published: (2026)

Aligning Actions and Walking to LLM-Generated Textual Descriptions
by: Chivereanu, Radu, et al.
Published: (2024)

Gait Recognition from Highly Compressed Videos
by: Niculae, Andrei, et al.
Published: (2024)

RoCode: A Dataset for Measuring Code Intelligence from Problem Definitions in Romanian
by: Cosma, Adrian, et al.
Published: (2024)

From Language Models over Tokens to Language Models over Characters
by: Vieira, Tim, et al.
Published: (2024)

Evaluating Character Understanding of Large Language Models via Character Profiling from Fictional Works
by: Yuan, Xinfeng, et al.
Published: (2024)

Word Recovery in Large Language Models Enables Character-Level Tokenization Robustness
by: Yang, Zhipeng, et al.
Published: (2026)

Enhancing Character-Level Understanding in LLMs through Token Internal Structure Learning
by: Xu, Zhu, et al.
Published: (2024)

Large Language Models Lack Understanding of Character Composition of Words
by: Shin, Andrew, et al.
Published: (2024)

Empowering Character-level Text Infilling by Eliminating Sub-Tokens
by: Ren, Houxing, et al.
Published: (2024)

Improving Language and Modality Transfer in Translation by Character-level Modeling
by: Tsiamas, Ioannis, et al.
Published: (2025)

Spelling-out is not Straightforward: LLMs' Capability of Tokenization from Token to Characters
by: Hiraoka, Tatsuya, et al.
Published: (2025)

Gumbel Machine: Counterfactual Student Writing Generation via Gumbel Noise Steering
by: McNichols, Hunter, et al.
Published: (2026)

Understanding the Emergence of Seemingly Useless Features in Next-Token Predictors
by: Rofin, Mark, et al.
Published: (2026)

CharacterBench: Benchmarking Character Customization of Large Language Models
by: Zhou, Jinfeng, et al.
Published: (2024)

Training a Bilingual Language Model by Mapping Tokens onto a Shared Character Space
by: Rom, Aviad, et al.
Published: (2024)

From Characters to Tokens: Dynamic Grouping with Hierarchical BPE
by: Dolga, Rares, et al.
Published: (2025)

OpenLLM-Ro -- Technical Report on Open-source Romanian LLMs
by: Masala, Mihai, et al.
Published: (2024)

Evaluating Language Model Character Traits
by: Ward, Francis Rhys, et al.
Published: (2024)

Dialectal and Low-Resource Machine Translation for Aromanian
by: Jerpelea, Alexandru-Iulius, et al.
Published: (2024)

CharBench: Evaluating the Role of Tokenization in Character-Level Tasks
by: Uzan, Omri, et al.
Published: (2025)

Revisiting Character-level Adversarial Attacks for Language Models
by: Rocamora, Elias Abad, et al.
Published: (2024)

Understanding and Mitigating Tokenization Bias in Language Models
by: Phan, Buu, et al.
Published: (2024)

From Tokens To Agents: A Researcher's Guide To Understanding Large Language Models
by: Barolo, Daniele
Published: (2026)