TL;DR
Aslema, the system for NADI 2026 Shared Task 5, which consists of two subtasks: intent recognition and slot filling and four omni LLMs in a zero-shot setting are presented and it is shown that fine-tuning consistently out-performs zero-shot inference.
This work evaluates four omni LLMs in a zero-shot setting and shows that fine-tuning consistently outperforms zero-shot inference, and explores synthetic data augmentation by using an LLM to generate culturally grounded Tunisian Derja utterances, followed by voice cloning to generate synthetic speech.
This paper details the Eloquence team's approach to Task 2 of the 2nd MLC-SLM challenge at Interspeech 2026, which involves multilingual Multiple-Choice Question Answering (MCQA) across 21 languages. Three approaches are explored. First, we fine-tune Voxtral-Mini-3B via LoRA with cross-lingual data augmentation, ASR tr...
Results show that the two subtasks behave differently in practice: satirical headline detection is better handled by prompting a large instruction-following model, while LLM-generated joke detection benefits more from supervised learning over dataset-specific stylistic and length-related cues.
The results show that combining a large zero-shot model with a smaller adapted model provides a practical framework for multilingual and fine-grained hallucination localization, while also highlighting the difficulty of transferring development-set improvements to hidden test data.
It is found that fine-tuning is a prerequisite for these inference-time techniques: without domain adaptation, prompting and shallow fusion cause severe hallucination on air traffic control speech, suggesting that domain difficulty determines whether prompting can work without fine-tuning.
The system is a single 2B vision-language model that answers multiple-choice questions about ten-minute egocentric videos in one greedy forward pass, and reaches 89% of the accuracy of the large agentic pipeline using 1.1% of its parameters.
We use cookies to run the site and, with your consent, for analytics and to show ads. See our Cookie Policy.
