Building a 100-question stereonet eval to test AI models' geology and 3D spatial reasoning
DynamicWebPaige · x · 2026-11-17
Inspired by conversations with Rice University earth science faculty, DynamicWebPaige is building a 100-question stereonet eval testing whether AI models genuinely understand geophysics and 3D spatial reasoning.
Sample questions require converting plane orientations to Right-Hand Rule strike/dip notation and deriving lower-hemisphere pole plunge/trend via spherical trigonometry, with structured JSON ground truth for graders — a hardcore test of spatial reasoning beyond typical benchmarks.
More from Research
- Sharp minima debate continues: Hessian still seems to help for LLMs in practice — burny_tech · 2026-11-17
- Bengio team's 2017 paper resurfaces: sharp minima can generalize for deep nets — burny_tech · 2026-11-17
- LLMs 101: free 2026 guide covers tokens to local deployment from first principles — CatAstro_Piyush · 2026-11-17
- Resurfaced Bengio paper: sharp minima can generalize for deep nets — burny_tech · 2026-11-17
- NUS Computing names nine inaugural Ascent Fellows for early-career researchers — rishabh16_ · 2026-11-17
- RLAD Paper: Two-Player RL Teaches LLMs to Discover Reasoning Abstractions — burny_tech · 2026-11-17