Reproduce “When Summaries Distort Decisions: Information Fidelity in LLM-Compressed Financial Analysis” on OpenResearch
Turn your favorite coding agents into research agents that review literature, develop hypotheses, and run experiments.
Universal app for Apple silicon and Intel. Requires macOS 11 or later.
About this paper
Hoyoung Lee, Suhwan Park, Seunghan Lee, Jun Seo, Jaehoon Lee, Sungdong Yoo, Minjae Kim, CheolWon Na and 10 more
The authors test whether compressed financial disclosures preserve source-implied decisions, measuring shifts in bear, neutral, and bull judgments by two decision models. Agentic Context Compression (ACC) compares contextualized summaries, audits disagreements against source snippets, and selects one intact. At a 20-bullet budget, ACC reduced MD&A flips from 31.5% to 19.5%; flips remained, and the results do not validate investment judgments.
Summary by alphaXiv.
Tip: change arxiv.org to autoarxiv.org in any arXiv abstract link to open the paper here.