Forecastbench is a dynamic, contamination-free benchmark of LLM forecasting accuracy with human comparison groups, serving as a valuable proxy for general intelligence.
-
Updated
Sep 10, 2026 - Python
#
forecastbench
Here are 5 public repositories matching this topic...
Local Qwen 3.5 4B yes/no forecaster. Brier 0.186 on 1,662 held-out ForecastBench questions. No API key, no cloud.
calibration forecasting gradio prediction-markets llm local-inference llama-cpp qwen brier-score forecastbench
-
Updated
Apr 18, 2026 - Python
Reproducible LLM evaluation of ForecastBench rationales, with dual blinded scoring, human validation, accuracy analysis, and an interactive dashboard
-
Updated
Sep 1, 2026 - HTML
Co-evolving Synthetic Intelligence Systems (SIS) with dignity, rhythm, and sacred potential.
forecasting ocap longevity fpic green-ai doughnut-economics ai-governance care-principles indigenous-data-sovereignty human-enhanced-performance synthetic-intelligence-systems forecastbench matakitenga story-soil
-
Updated
Jun 13, 2026
A verified map of the AI-forecasting research area: ForecastBench, OpenForecaster, and FutureX — what each measures, what their artifacts really contain, which numbers are current, and where the open problems are.
calibration benchmarks prediction-markets llm-evaluation ai-forecasting forecastbench superforecasting
-
Updated
Aug 22, 2026 - Python
Add this topic to your repo
To associate your repository with the forecastbench topic, visit your repo's landing page and select "manage topics."