paper-with-me

홈 › Papers

You Are What You Train: Effects of Data Composition on Training Context-aware Machine Translation Models

2025-09-17 · Paweł Mąka, Yusuf Can Semerci, Jan Scholtes, Gerasimos Spanakis arxiv

Achieving human-level translations requires leveraging context to ensure coherence and handle complex phenomena like pronoun disambiguation. Sparsity of contextually rich examples in the standard training data has been hypothesized as the reason for the difficulty of context utilization. In this work, we systematically validate this claim in both single- and multilingual settings by constructing training datasets with a controlled proportions of contextually relevant examples. We demonstrate a strong association between training data sparsity and model performance confirming sparsity as a key bottleneck. Importantly, we reveal that improvements in one contextual phenomenon do no generalize to others. While we observe some cross-lingual transfer, it is not significantly higher between languages within the same sub-family. Finally, we propose and empirically evaluate two training strategies designed to leverage the available data. These strategies improve context utilization, resulting in accuracy gains of up to 6 and 8 percentage points on the ctxPro evaluation in single- and multilingual settings respectively.

📄 PDF Abstract BibTeX arXiv:2509.14031

Code (0)

등록된 구현이 없습니다.

Tasks

Cross-Lingual TransferMachine Translation

Similar Papers 제목 키워드 기반

Neural Decomposition: Functional ANOVA with Variational Autoencoders

2020-06-25 · Kaspar Märtens, Christopher Yau

Variational Autoencoders (VAEs) have become a popular approach for dimensionality reduction. However, despite their ability to identify latent low-dimensional structures embedded within high-dimensional data, these laten…

DecoderDimensionality Reduction

What Training Data Teaches RL Memory Agents: An Empirical Study of Curriculum Effects in Memory-Augmented QA

2026-05-21 · Xinjie He, Zhiyuan Lin, Su Liu, Jialun Wu 외 arxiv

Reinforcement learning (RL) has emerged as a viable recipe for training LLM agents to reason over external memory banks in multi-session dialogue. Existing work trains exclusively on a single benchmark, leaving open how …

Reinforcement Learning

How Does RL Post-training Induce Skill Composition? A Case Study on Countdown

2025-12-01 · Simon Park, Simran Kaur, Sanjeev Arora arxiv

While reinforcement learning (RL) successfully enhances reasoning in large language models, its role in fostering compositional generalization (the ability to synthesize novel skills from known components) is often confl…

Reinforcement Learning

What Meaning-Form Correlation Has to Compose With

2020-12-07 · Timothee Mickus, Timothée Bernard, Denis Paperno

Compositionality is a widely discussed property of natural languages, although its exact definition has been elusive. We focus on the proposal that compositionality can be assessed by measuring meaning-form correlation. …

Form

What Meaning-Form Correlation Has to Compose With: A Study of MFC on Artificial and Natural Language

2020-12-01 · COLING 2020 8 · Timothee Mickus, Timoth{\'e}e Bernard, Denis Paperno

Compositionality is a widely discussed property of natural languages, although its exact definition has been elusive. We focus on the proposal that compositionality can be assessed by measuring meaning-form correlation. …

Form