Extractive Question Answering for Causal Relations on FinCausal Spanish 2026 (test)
4.7753LLM Judge ScoreHSA_CORAL
Evaluation Results
| Method | Links | |
|---|---|---|
| HSA_CORALBase Model=GPT-4.1 mini, Fine-tuned=yes, Corpus=Multilingual (en+es)2026.06 | 4.7753 | |
| GPT-4.1 miniFine-tuned=yes, Corpus=Monolingual (es)2026.06 | 4.7674 | |
| GPT-5.2Fine-tuned=no, Corpus=-2026.06 | 4.735 | |
| GPT-4.1 miniFine-tuned=yes, Corpus=Monolingual (en)2026.06 | 4.7141 | |
| GPT-3.5 turboFine-tuned=no, Corpus=-2026.06 | 4.706 | |
| BART Facebook BaseFine-tuned=yes, Corpus=Multilingual (en+es)2026.06 | 4.03 | |
| BERT Base MultilingualFine-tuned=yes, Corpus=Multilingual (en+es)2026.06 | 3.981 | |
| Llama-3.1 8BFine-tuned=yes, Corpus=Multilingual (en+es)2026.06 | 3.91 |