EACL 2026: Accepted Papers

Three Papers Accepted to EACL 2026

We are pleased to announce that three papers authored by members of our lab have been accepted to the European Chapter of the Association for Computational Linguistics (EACL) 2026.

---

Knowledge Augmentation Enhances Token Classification for Recipe Understanding

Nuhu Ibrahim, Robert Stevens, Riza Batista-Navarro

This paper investigates how external knowledge sources can be leveraged to improve token-level classification for recipe understanding tasks. The results demonstrate that knowledge augmentation significantly enhances model robustness and interpretability in procedural text analysis.

---

Lost in Formatting: How Output Formats Skew LLM Performance on Information Extraction

Rishi Ravikumar, Nuhu Ibrahim, Riza Batista-Navarro

This work examines how different output formatting choices can significantly impact large language model performance on information extraction tasks, highlighting an often-overlooked source of evaluation bias in large language model evaluation.

---

ReciFine: Finely Annotated Recipe Dataset for Controllable Recipe Generation

Nuhu Ibrahim, Rishi Ravikumar, Robert Stevens, Riza Batista-Navarro

This paper introduces ReciFine, a richly annotated recipe dataset designed to support controllable recipe generation. The dataset provides a valuable resource for future research in culinary NLP and text generation.

---

We congratulate all authors and look forward to presenting this work at EACL 2026. Additional details, including preprints, data, code, and presentation information, will be shared in due course.

We hope to see you in Rabat in March!