A bilingual benchmark for evaluating large language models on buy-side equity research and valuation tasks.
BuySideFinBench targets the analytical reasoning that distinguishes a buy-side equity research analyst from a generalist financial reader. Unlike most finance LLM benchmarks that focus on sell-side / news-driven tasks (sentiment, summarization, headline interpretation) or surface-level multiple-choice knowledge, BuySideFinBench… See the full description on the dataset page:
https://huggingface.co/datasets/cindy90/BuySideFinBench.
