What if physics isn't fine-tuned, just well-shaped?

GIFT derives Standard Model parameters from the geometry of a single 7-dimensional manifold. No free parameters. No fitting. Every prediction is a consequence of shape: E₈×E₈ gauge theory compactified on a G₂-holonomy manifold K₇ with Betti numbers (b₂, b₃) = (21, 77).

All posts on giftheory.substack.com.

Dimensional reduction: E₈×E₈ (496D) → AdS₄ × K₇ (11D) → Standard Model (4D)

Details: Statistical Evidence

The framework contains no continuous adjustable parameters fitted to data. However, it makes discrete structural choices: E₈×E₈ as gauge group, K₇ with (b₂=21, b₃=77), TCS building blocks. These are mathematically motivated but constitute model selection. The framework predicts observables given these choices: it does not explain why nature chose this geometry.

Statistical validation shows (b₂=21, b₃=77) is the unique optimum among 3,070,396 tested configurations. This doesn't explain the choice, but establishes it is not arbitrary.

@software{gift_framework,
  title   = {GIFT: Geometric Information Field Theory},
  author  = {de La Fournière, Brieuc},
  year    = {2026},
  url     = {https://github.com/gift-framework/GIFT},
  version = {3.4}
}

See CITATION.md for additional formats.

MIT License, see LICENSE

Gift from bit