Papers by Natalie Schluter

15 papers
The glass ceiling in NLP (D18-1)

Copied to clipboard

Challenge: a glass ceiling exists within the field of NLP, but no study has examined this issue . female representation in Computer Science is lower than the average STEM field .
Approach: They propose to use a mathematical model to show that a glass ceiling exists in NLP . they find that there is a growing mentor gender gap and a disparity between mentors .
Outcome: The proposed model shows that a glass ceiling exists within the field of NLP since the mid 2000s.
The Role of Prosody in Spoken Question Answering (2025.findings-naacl)

Copied to clipboard

Challenge: lexical information is not available in most models, but prosody is important in understanding spoken language.
Approach: They investigate the role of prosody in the process of answering a spoken question by isolating prosodic and lexical information from a natural speech dataset.
Outcome: The proposed models can perform reasonably well on the SLUE-SQA-5 dataset, but when lexical information is available, models tend to predominantly rely on it.
Akan Cinematic Emotions (ACE): A Multimodal Multi-party Dataset for Emotion Recognition in Movie Dialogues (2025.findings-acl)

Copied to clipboard

Challenge: Akan Cinematic Emotions (AkaCE) is the first multimodal emotion dialogue dataset for an African language . it contains 385 emotion-labeled dialogues and 6162 utterances across audio, visual, and textual modalities, along with word-level prosodic prominence annotations.
Approach: They propose to use AkaCE to analyze African cinematic emotions using word-level prosodic prominence annotations.
Outcome: The Akan Cinematic Emotions (AkaCE) dataset addresses the significant lack of resources for low-resource languages in emotion recognition research.
Discriminating Form and Meaning in Multilingual Models with Minimal-Pair ABX Tasks (2025.emnlp-main)

Copied to clipboard

Challenge: Existing studies have shown that multilingual models encode languagespecific information and language-agnostic features, but the nature and interaction of these representations is not fully understood.
Approach: They propose a set of training-free ABX-style discrimination tasks to evaluate how multilingual language models represent language identity (form) and semantic content (meaning).
Outcome: The proposed tasks show that language discrimination declines over training and strengthens over time and stabilizes in deeper layers.
DaNewsroom: A Large-scale Danish Summarisation Dataset (2020.lrec-1)

Copied to clipboard

Challenge: Existing datasets for automatic summarisation are English-oriented . however, only very limited datasets exist in languages other than English .
Approach: They present the first large-scale non-English dataset specifically curated for automatic summarisation.
Outcome: The proposed dataset is the first for the Danish language and is compared with existing datasets.
The Word Analogy Testing Caveat (N18-2)

Copied to clipboard

Challenge: a number of word analogy tests are used to evaluate word embeddings . word embeds are used as a proxy for semantics and syntax à la Harris .
Approach: They propose to use word embeddings as a proxy for distributional similarity . they propose to apply a transfer learning approach to word embeds to improve performance .
Outcome: The proposed method improves performance across a wide range of NLP tasks.
Construction of Paired Knowledge Graph - Text Datasets Informed by Cyclic Evaluation (2024.lrec-main)

Copied to clipboard

Challenge: Prior studies have shown that sequence-to-sequence models learn to hallucinate when the conditioning data has poor correlation with the sequence being produced.
Approach: They construct a dataset that pairs Knowledge Graphs (KG) and text together and compare their results to a cyclic evaluation model.
Outcome: The proposed model performs better on cyclic generation of KGs than on KG-T, but less well on synchronization of KTs.
Does Context Matter? A Prosodic Comparison of English and Spanish in Monolingual and Multilingual Discourse Settings (2025.emnlp-main)

Copied to clipboard

Challenge: a large number of studies on prosody in languages have focused on monolingual discourse contexts . a recent study focused on the prosodic features of monolingual speech in multilingual contexts.
Approach: They compare prosody of monolingual English and Spanish in monolingual and multilingual settings . they find that monolingual speech produced in a monolingual context is prosodically different from that produced in multilingual context .
Outcome: The proposed study is the first to incorporate multilingual discourse contexts into the study of native-level monolingual prosody.
When data permutations are pathological: the case of neural natural language inference (D18-1)

Copied to clipboard

Challenge: Recent research indicates that these datasets are trivially annotated to a large degree.
Approach: They propose to use a trending NLP task to test a neural architecture for NLI . they show a pathological sensitivity of the systems to permutations of the training set .
Outcome: The proposed model outperforms the baseline model by multiple percentage points . the proposed model is based on the proposed dataset and the soundness of the task .
Training Bilingual LMs with Data Constraints in the Targeted Language (2025.findings-acl)

Copied to clipboard

Challenge: a large number of languages have insufficient data for pretraining, but most non-English models are trained on scrapes of the web.
Approach: They propose to use data from an auxiliary language to boost model performance . they quantify the performance gap between training with data in a data-rich auxiliary and training in the target language .
Outcome: The proposed method boosts model performance in a target language with insufficient data . it also explores the benefits of translation systems and the limitations of model scaling when data is limited.
How Value Induction Reshapes LLM Behavior (2026.findings-acl)

Copied to clipboard

Challenge: Induction of values into LLMs can have unintended effects on the user interacting with it.
Approach: They investigate the unintended effects of value incorporation into models by fine-tuning existing preference datasets and measuring their effect on safety, anthropomorphism and QA benchmarks.
Outcome: The proposed model improves safety, anthropomorphism and QA benchmarks by inducing values and incorporating values into the model.
GrammaMT: Improving Machine Translation with Grammar-Informed In-Context Learning (2025.acl-long)

Copied to clipboard

Challenge: Experiments show that GrammaMT enhances translation performance on open-source instruction-tuned LLMs for various low- to high-resource languages across three benchmarks: (1) largest corpus, (2) challenging 2023 SIGMORPHON Shared Task data, (3) even in an out-of-domain setting with FLORES.
Approach: They propose a grammatically-aware prompting approach that uses Interlinear Glossed Text . they propose gloss-shot, chain-gloss and model-glooss prompting strategies that are training-free .
Outcome: Experiments show that GrammaMT improves translation performance on open-source instruction-tuned LLMs for various low- to high-resource languages across three benchmarks.
MassiveSumm: a very large-scale, very multilingual, news summarisation dataset (2021.emnlp-main)

Copied to clipboard

Challenge: Current research in automatic summarisation is expensive to create, posing a challenge for any language.
Approach: They propose to use a large-scale multilingual summarisation dataset with articles in 92 languages and more than 35 writing scripts to generate a multilingual dataset.
Outcome: The proposed method is the largest, most inclusive, existing dataset and one of the largest and most inclusive datasets for any NLP task.
Baselines and Test Data for Cross-Lingual Inference (L18-1)

Copied to clipboard

Challenge: Recent research on textual entailment is limited to English, but it is expanding to other languages.
Approach: They propose to extend the research in SNLI-style natural language inference toward multilingual evaluation by using cross-lingual word embeddings and machine translation.
Outcome: The proposed system scores an average accuracy of just over 75%, but it is not perfect.
Recurrent models and lower bounds for projective syntactic decoding (N19-1)

Copied to clipboard

Challenge: a string of recent work has attempted to delve into the formal properties of neural network topology choices.
Approach: They propose to use recurrent models to perform projective maximum spanning tree decoding . they also prove the lower bounds of projective maximal spanning trees .
Outcome: The proposed model can perform better than Eisner's model, proving it impossible to predict a projective MST.

What is GenGO?

GenGO is an NLP powered publication search system. It currenctly indexes 30k+ papers from ACL Anthology, and implements multi-aspect summarization, semantic search, and more!

Information

About
Limitations