Multi-Doc-2025: Financial QA Benchmark from S&P 500 SEC 10-K Reports

Multi-Doc-2025 is a financial question-answering benchmark built from SEC Form 10-K annual reports of S&P 500 companies. The dataset is designed to evaluate retrieval-augmented generation and financial QA systems under cross-company, cross-year, and hybrid-modal reasoning settings. It was created by Anonymous-Team-HC-RAG and last updated on 2026-05-27.

Use Cases

• Benchmarking cross-company reasoning systems based on questions spanning multiple S&P 500 firms
• Evaluating cross-year financial reasoning models using data from different annual reports
• Testing hybrid-modal QA systems that must process both text and tables from SEC filings

Strengths

• Focuses on three reasoning settings not jointly covered by existing financial QA benchmarks
• Built from authoritative SEC Form 10-K annual reports of S&P 500 companies
• Designed specifically for evaluating retrieval-augmented generation systems

Limitations

• Column-level documentation is absent; field semantics must be inferred after download
• Row count is unknown, which may limit suitability assessment
• The dataset description is incomplete and refers to an external page for full details

Provenance

Source

SEC Form 10-K annual reports of S&P 500 companies

Collection Method

Likely extracted and processed to create a question-answering benchmark

Freshness

Last updated 2026-05-27 11:49:09