ALEE generates AMR-based English minimal pairs with fine-grained semantic shifts, translates them, and evaluates embedding models on 275+ languages to expose cross-lingual gaps linked to training data and tokenization.
hub
How Good is Your Tokenizer? On the Monolingual Performance of Multilingual Language Models , booktitle =
21 Pith papers cite this work, alongside 21 external citations. Polarity classification is still indexing.
21
Pith papers citing it
21
external citations · OpenAlex
hub tools
representative citing papers
LangMAP adapts UnigramLM for multilingual use to deliver language-specific tokenization from a shared vocabulary, boosting boundary alignment metrics across natural and programming languages with mixed downstream fine-tuning gains.
Evaluation across 1.1 million instances shows sycophancy rates spike in low-resource languages, remain topic-agnostic, and correlate with tokenizer fertility.
UA-Legal-Bench is a new five-task benchmark for Ukrainian legal reasoning that demonstrates task-dependent few-shot prompting effects and the need for macro-F1 over accuracy on imbalanced classes.
Training-language dominance, not English inherent properties, determines brain-LLM alignment across English, Chinese, and French, with additional independent effects from typological distance concentrated in syntactic brain regions.
ToaST uses vocabulary-independent split trees and integer programming to produce tokenizers with over 11% fewer tokens than BPE, WordPiece, and UnigramLM while improving 1.5B-parameter LM scores on CORE.
OPT releases open decoder-only transformers up to 175B parameters that match GPT-3 performance at one-seventh the carbon cost, along with code and training logs.
On chemistry SMILES with a fixed 165-token base, BPE and Unigram-LM produce near-disjoint vocabularies (Jaccard ≤0.161) and Unigram-LM emits 29–41% more tokens across 22 matched conditions.
A Bayesian framework decomposes mLLM variance, showing language features explain 79-92% of language identity variance and that model identity vs. benchmark-model interactions dominate differently for understanding versus reasoning tasks.
Vocabulary adaptation via targeted token addition and replacement improves semantic similarity, domain word usage, and training efficiency for LLM summarization in legal and medical domains.
In compute-optimal regimes, language model parameter count scales proportionally with data bytes rather than tokens, and the optimal compression rate decreases with increasing compute.
BLOOM is a 176B-parameter open-access multilingual language model trained on the ROOTS corpus that achieves competitive performance on benchmarks, with improved results after multitask prompted finetuning.
TOTEN is a knowledge-based system for structure-preserving representation of physical quantities and technical notation in Brazilian Portuguese using an ontology of engineering entities and external authorities, outperforming statistical baselines in atomicity and reconstruction.
A 355M-parameter byte-level LM on 80B multilingual tokens exhibits UTF-8 validity converging after 4.2B tokens versus 2.1B for perplexity, with higher validity on rare characters than common ones.
Pilot evaluation of language-specific versus multilingual LoRA adapters on Qwen2.5-VL-3B for curator-guided BLV art descriptions in three languages.
Per-language architecture selection among generalists, specialists, and ensembles achieves 0.796 macro F1 across 22 languages in SemEval-2026 Task 9.
StarCoderBase matches or beats OpenAI's code-cushman-001 on multi-language code benchmarks; the Python-fine-tuned StarCoder reaches 40% pass@1 on HumanEval while retaining other-language performance.
Replacing tokens, freezing the corresponding embeddings, and tuning the rest of the model improves NLU performance on low-resource languages compared to full fine-tuning.
HAMR combines bi-level meta-learned instance reweighting with KNN-based neighborhood resampling to improve NLP performance under class imbalance, evaluated on six NER and classification benchmarks.
A heterogeneous ensemble of XLM-RoBERTa-large and mDeBERTa-v3-base with independent task modeling and class weighting is reported as effective for multilingual, multicultural, and multievent online polarization detection.
citing papers explorer
-
ALEE: Any-Language Evaluation of Embeddings via English-Centric Minimal Pairs
ALEE generates AMR-based English minimal pairs with fine-grained semantic shifts, translates them, and evaluates embedding models on 275+ languages to expose cross-lingual gaps linked to training data and tokenization.
-
LangMAP: A Language-Adaptive Approach to Tokenization
LangMAP adapts UnigramLM for multilingual use to deliver language-specific tokenization from a shared vocabulary, boosting boundary alignment metrics across natural and programming languages with mixed downstream fine-tuning gains.
-
Sycophancy as a Multilingual Alignment Failure: How Safety Degrades Across Languages, Topics, and Models
Evaluation across 1.1 million instances shows sycophancy rates spike in low-resource languages, remain topic-agnostic, and correlate with tokenizer fertility.
-
UA-Legal-Bench: A Benchmark for Evaluating Large Language Models on Ukrainian Legal Reasoning
UA-Legal-Bench is a new five-task benchmark for Ukrainian legal reasoning that demonstrates task-dependent few-shot prompting effects and the need for macro-F1 over accuracy on imbalanced classes.
-
Brain-LLM Alignment Tracks Training Data, Not Typology
Training-language dominance, not English inherent properties, determines brain-LLM alignment across English, Chinese, and French, with additional independent effects from typological distance concentrated in syntactic brain regions.
-
Tokenization with Split Trees
ToaST uses vocabulary-independent split trees and integer programming to produce tokenizers with over 11% fewer tokens than BPE, WordPiece, and UnigramLM while improving 1.5B-parameter LM scores on CORE.
-
OPT: Open Pre-trained Transformer Language Models
OPT releases open decoder-only transformers up to 175B parameters that match GPT-3 performance at one-seventh the carbon cost, along with code and training logs.
-
Where to cut, how deep: BPE and Unigram-LM on chemistry SMILES
On chemistry SMILES with a fixed 165-token base, BPE and Unigram-LM produce near-disjoint vocabularies (Jaccard ≤0.161) and Unigram-LM emits 29–41% more tokens across 22 matched conditions.
-
DEPART: DEcomposing PARiTy across Multilingual LLMs
A Bayesian framework decomposes mLLM variance, showing language features explain 79-92% of language identity variance and that model identity vs. benchmark-model interactions dominate differently for understanding versus reasoning tasks.
-
Learning Faster with Better Tokens: Parameter-Efficient Vocabulary Adaptation for Specialized Text Summarization
Vocabulary adaptation via targeted token addition and replacement improves semantic similarity, domain word usage, and training efficiency for LLM summarization in legal and medical domains.
-
Compute Optimal Tokenization
In compute-optimal regimes, language model parameter count scales proportionally with data bytes rather than tokens, and the optimal compression rate decreases with increasing compute.
-
BLOOM: A 176B-Parameter Open-Access Multilingual Language Model
BLOOM is a 176B-parameter open-access multilingual language model trained on the ROOTS corpus that achieves competitive performance on benchmarks, with improved results after multitask prompted finetuning.
-
Toten: A Knowledge-Based System For Structure-Preserving Representation Of Physical Quantities And Technical Notation In Brazilian Portuguese
TOTEN is a knowledge-based system for structure-preserving representation of physical quantities and technical notation in Brazilian Portuguese using an ontology of engineering entities and external authorities, outperforming statistical baselines in atomicity and reconstruction.
-
Beyond Perplexity: UTF-8 Validity in Byte-aware Language Models
A 355M-parameter byte-level LM on 80B multilingual tokens exhibits UTF-8 validity converging after 4.2B tokens versus 2.1B for perplexity, with higher validity on rare characters than common ones.
-
A Pilot Study on Curator-Guided Multilingual Art Description for Blind and Low-Vision Audiences with Small Vision-Language Models
Pilot evaluation of language-specific versus multilingual LoRA adapters on Qwen2.5-VL-3B for curator-guided BLV art descriptions in three languages.
-
MKJ at SemEval-2026 Task 9: A Comparative Study of Generalist, Specialist, and Ensemble Strategies for Multilingual Polarization
Per-language architecture selection among generalists, specialists, and ensembles achieves 0.796 macro F1 across 22 languages in SemEval-2026 Task 9.
-
StarCoder: may the source be with you!
StarCoderBase matches or beats OpenAI's code-cushman-001 on multi-language code benchmarks; the Python-fine-tuned StarCoder reaches 40% pass@1 on HumanEval while retaining other-language performance.
-
Modular Monolingual Adaptation using Pretrained Language Models
Replacing tokens, freezing the corresponding embeddings, and tuning the rest of the model improves NLU performance on low-resource languages compared to full fine-tuning.
-
Syntax as a Rosetta Stone: Universal Dependencies for In-Context Coptic Translation
HAMR combines bi-level meta-learned instance reweighting with KNN-based neighborhood resampling to improve NLP performance under class imbalance, evaluated on six NER and classification benchmarks.
-
YEZE at SemEval-2026 Task 9: Detecting Multilingual, Multicultural and Multievent Online Polarization via Heterogeneous Ensembling
A heterogeneous ensemble of XLM-RoBERTa-large and mDeBERTa-v3-base with independent task modeling and class weighting is reported as effective for multilingual, multicultural, and multievent online polarization detection.
- MinGram: A Minimalist Unigram Tokenizer with High Compression and Competitive Morphological Alignment