ReML-AI

[AAAI-26] Reasoning Transfer for an Extremely Low-Resource and Endangered Language: Bridging Languages Through Sample-Efficient Language Understanding

[EACL 2026] LaCoMSA: Language-Consistency Multilingual Self-Alignment with Latent Representation Rewarding

[ACL 2026] Disentangling Continued Pre-training: Attention-Driven Routing and Semantic Hub Preservation in Language Adaptation

[LREC 2026] VIVID: A Culturally Grounded Benchmark Exposing the Figurative Language Gap in Vietnamese NLP

[KDD 2026] IRLBench: A Multi-modal, Culturally Grounded, Parallel Irish-English Benchmark for Open-Ended LLM Reasoning Evaluation

[EMNLP2025] Disentangling Language Understanding and Reasoning Structures in Cross-lingual Chain-of-Thought Prompting

[LREC 2026] Irish-BLiMP: A Linguistic Benchmark for Evaluating Human and Language Model Performance in a Low-Resource Setting