Sep 24, 2026 · Artificial Intelligence

Seed-2.1-Pro 0915 Stress Test: Honesty Under Tool Failures & Missing Data

The author evaluates Seed-2.1-Pro 0915 against its predecessor using simulated tool failures, cropped financial PDFs, and a full analysis pipeline, finding both models honest when evidence is missing but revealing post-processing unit conversion errors, concluding that model execution requires system-level verification.

LLM evaluationSeed-2.1-Proagent honesty

0 likes · 15 min read

Seed-2.1-Pro 0915 Stress Test: Honesty Under Tool Failures & Missing Data

AI Code to Success

Sep 18, 2026 · Artificial Intelligence

Doubao-Seed-2.1-pro Reads 232-Page Annual Reports: Building a Verified Financial Research Pipeline

The author tests Doubao-Seed-2.1-pro's multimodal and deep research capabilities by extracting structured financial data from 232-page annual reports, building a pipeline with cross-validation, Excel formulas, and HTML dashboards, then generalizing to BYD and Moutai, revealing that context understanding and evidence chains matter more than OCR accuracy.

AI in financeDoubao-Seed-2.1-proannual report processing

0 likes · 16 min read

Doubao-Seed-2.1-pro Reads 232-Page Annual Reports: Building a Verified Financial Research Pipeline

financial document analysis

Seed-2.1-Pro 0915 Stress Test: Honesty Under Tool Failures & Missing Data

Doubao-Seed-2.1-pro Reads 232-Page Annual Reports: Building a Verified Financial Research Pipeline