# UniBuc at SemEval-2024 Task 2: Tailored Prompting with Solar for Clinical NLI

arXiv: https://arxiv.org/abs/2610.00408 · HTML: https://mlsanity.com/paper/2610.00408/

## Key facts

- No citing papers in the map yet (data through 14 Oct 2026).
- Builds on 11 of its 19 references that are in the map.
- First posted on arXiv 30 Sep 2026.

## Builds on (most cited first, 11 of 11)

- [Llama 2: Open Foundation and Fine-Tuned Chat Models](https://mlsanity.com/paper/2307.09288/) (2023)
- [Unsupervised Cross-lingual Representation Learning at Scale](https://mlsanity.com/paper/1911.02116/) (2019)
- [DeBERTa: Decoding-enhanced BERT with Disentangled Attention](https://mlsanity.com/paper/2006.03654/) (2020)
- [Domain-Specific Language Model Pretraining for Biomedical Natural Language Processing](https://mlsanity.com/paper/2007.15779/) (2020)
- [How to Fine-Tune BERT for Text Classification?](https://mlsanity.com/paper/1905.05583/) (2019)
- [SOLAR 10.7B: Scaling Large Language Models with Simple yet Effective Depth Up-Scaling](https://mlsanity.com/paper/2312.15166/) (2023)
- [Lessons from Natural Language Inference in the Clinical Domain](https://mlsanity.com/paper/1808.06752/) (2018)
- [A Comprehensive Evaluation of Large Language Models on Benchmark Biomedical Text Processing Tasks](https://mlsanity.com/paper/2310.04270/) (2023)
- [SemEval-2024 Task 2: Safe Biomedical Natural Language Inference for Clinical Trials](https://mlsanity.com/paper/2404.04963/) (2024)
- [NLI4CT: Multi-Evidence Natural Language Inference for Clinical Trial Reports](https://mlsanity.com/paper/2305.03598/) (2023)
- [CHARD: Clinical Health-Aware Reasoning Across Dimensions for Text Generation Models](https://mlsanity.com/paper/2210.04191/) (2022)

## Abstract (arXiv)

This paper describes the approach of the UniBuc team in tackling the SemEval 2024 Task 2: Safe Biomedical Natural Language Inference for Clinical Trials. We used SOLAR Instruct, without any fine-tuning, while focusing on input manipulation and tailored prompting. By customizing prompts for individual CTR sections, in both zero-shot and few-shots settings, we managed to achieve a consistency score of 0.72, ranking 14th in the leaderboard. Our thorough error analysis revealed that our model has a tendency to take shortcuts and rely on simple heuristics, especially when dealing with semantic-preserving changes.