FINKRX: Establishing Best Practices for Korean Financial NLP

In this work, we present the first open leaderboard for evaluating Korean large language models focused on finance. Operated for about eight weeks, the leaderboard evaluated 1,119 submissions on a closed benchmark covering five MCQA categories: finance and accounting, stock price prediction, domestic company analysis, financial markets, and financial agent tasks and one open-ended qa task. Building on insights from these evaluations, we release an open instruction dataset of 80k instances and summarize widely used training strategies observed among top-performing models. Finally, we introduce FINKRX, a fully open and transparent LLM built using these best practices. We hope our contributions help advance the development of better and safer financial LLMs for Korean and other languages. 1 Category Examples Financial Markets 642 total 다음 중 대한민국 주식시장 매매 제도에 대한 기술로 알맞은 것은 무엇인가? Which of the following descriptions is correct regarding the trading system of the Korean stock market? A. Opening time is 10:00 AM. B. The daily price limit for the KOSPI market is ±15% of the previous day's closing price. [...] Finance and Accounting 1,450 total 다음 중 화폐의 시간가치에 관한 설명으로 옳지 않은 것은 무엇인가? Which of the following statements about the value of money is incorrect? A. In monthly compounding, the monthly interest rate is calculated by dividing the annual [...] B. Given the same initial investment and conditions, compound interest yields higher [...] Domestic Company Analysis 2,039 total 엑세스바이오의 COVID-19 진단 제품의 매출 기여와 미국 시장 판매에 대해서 올바른 것은? What is correct regarding the sales contribution of Access Bio's COVID-19 diagnostic products and their sales in the U.S. market? A. Access Bio's COVID diagnostic products were developed for general health screening [...] B. Access Bio's COVID diagnostic products have demonstrated effectiveness through [...] Financial Agent 46 total 데이터프레임의 '종가' 열의 평균 값을 계산하는 코드를 고르시오. Choose the code that calculates the average value of the 'Closing Price' column in the DataFrame. A. df['Close Price'].mean() B. df['Total Traded Quantity'].median() [.