Papers for

multilingual chatbot designers

Papers whose findings have a practical use for this group, as judged from the abstract. Open a paper to read what it means in practice.

NileChat model translates English into 13 Arabic dialects in dialogue

alkhder at AlexandriaX-2026: Context-Aware English-to-Dialectal Arabic Dialogue Translation with QLoRA-Fine-Tuned NileChat-3B

Abstract: This paper describes team Alkhder's submission to Subtask 1 of the AlexandriaX-2026 shared task, which asks systems to translate English dialogue turns into thirteen Arabic dialects. We fine-tune NileChat-3B-Base with QLoRA on the Alexandria parallel dialogue corpus, training one set of adapter weights for all dialects. Each turn is translated conditioned on the target country and dialect, the domain, the participant roles, the speaker-addressee gender direction, and the preceding dialogue context. At inference, conversations are translated autoregressively and the model is fed its own previous outputs back as context without using any gold target-side test information. Under the constrained track, using only Alexandria data, it trains in about one hour and translates the private test set in under two hours. The system reaches a country-macro-averaged spBLEU of 20.36 and chrF++ of 36.64 on both tracks.

Thu 15 OctComputation and Language
The gist
Translating English conversations into various Arabic dialects is hard because there are many forms that depend on country, speaker roles, and context. The authors created a system by fine-tuning an AI model called NileChat-3B using special training on a dataset of English-Arabic dialogues. Their system translates each turn in a conversation taking into account the dialect, speaker roles, and previous conversation lines. It works automatically without needing correct target answers during translation and runs quickly. The system scores well on measuring how closely its output matches human translations.
Open → 2610.22525v1

Healthcare language models miss varied cultural insights in Indian maternal care

Measuring Cultural Alignment Beyond the Average: A Framework for Evaluating Maternal-Health LLM Interactions in Indian Contexts

Abstract: Existing evaluation methods for healthcare LLMs primarily assess factual correctness,safety, and fluency, while providing limited insight into whether generated interactions reflect culturally situated healthcare reasoning. This limitation is particularly important in maternal health, where care decisions are shaped by social and relational norms. We introduce MH-INDIC, a culturally grounded evaluation framework for maternal-health interactions in urban and semi-urban North Indian contexts that operationalises cultural behaviour through ten dimensions of maternal-health reasoning. Using a 26-item survey administered to 102 pregnant and postpartum women from urban and semi-urban North India, we evaluate ten LLMs. We distinguish population level cultural alignment from profile-level behavioural variation. Although several models approximate the human population-level distribution, all evaluated systems exhibit substantially lower variation across demographic and household profiles than the human cohort, revealing a gap between aggregate alignment and profile-conditioned sensitivity. As a downstream application of MH-INDIC, we use the strongest-aligned proprietary and open-source models to generate culturally conditioned maternal-health dialogues under zero-shot, self-conditioned, and human-grounded prompting. Human-grounded conditioning produces stronger profile alignment and dialogue quality ratings, suggesting that measured cultural profiles can improve the cultural grounding of generated interactions

Thu 8 OctComputation and Language
The gist
Current AI systems that help with health advice usually check if the info they give is true and safe but don’t see if the advice fits local culture. This matters a lot in Indian pregnancy care, where local traditions shape health choices. The authors created a way to measure if AI advice matches the cultural beliefs of women in urban and semi-urban North India. They found that AI systems often give similar advice to all profiles, unlike real people who vary their opinions based on different backgrounds. Conditioning AI with specific cultural info helps make its advice more accurate and relatable.
Open → 2610.11586v1

Rosetta system improves English to dialectal Arabic dialogue translation

Rosetta at AlexandriaX-2026: LoRA-Adapted NileChat for Context-Aware Dialectal Arabic Dialogue Translation

Abstract: This paper describes the Rosetta system for Subtask 1 (Context-Aware English-to-Dialectal Arabic Dialogue Translation) of the AlexandriaX shared task, participating in both constrained and unconstrained tracks. The approach fine-tunes a LoRA adapter on NileChat-3B using structured system/user prompts that condition generation on dialect and dialogue context. For the unconstrained track, the adapter is additionally pretrained on MADAR and PADIC. Rosetta ranked 4th in the constrained track (spBLEU 26.10) and 5th in the unconstrained track (spBLEU 25.09). The experimental results demonstrate that external pretraining helps only two of thirteen dialects while slightly hurting overall performance, suggesting negative transfer.

Wed 9 SeptComputation and Language
The gist
Translating English conversations into Arabic dialects is tricky because each dialect can be very different. The authors created Rosetta, a system that fine-tunes a small, adaptable model called LoRA on an existing chat model to handle these dialects better by using context from the conversation. They also experimented with additional training on other dialect datasets. Their system performed well, placing near the top in a competition, but found that extra training sometimes hurt overall accuracy, showing the challenge of mixing dialects.
Open → 2609.10395v1