Similar Items: Learning More from Less: Exploiting Counterfactuals for Data-Efficient Chart Understanding
- Why Low-Resource NLP Needs More Than Cross-Lingual Transfer: Lessons Learned from Luxembourgish
- Structure Liberates: How Constrained Sensemaking Produces More Novel Research Output
- Repetition over Diversity: High-Signal Data Filtering for Sample-Efficient German Language Modeling
- ReLay: Personalized LLM-Generated Plain-Language Summaries for Better Understanding, but at What Cost?
- Efficient Pre-Training with Token Superposition
- Accurate and Efficient Statistical Testing for Word Semantic Breadth