park-2025-validation-models-wordnet
IN premise — summaries/2026/08/24/park-2024-categorical-hierarchical-concepts-s2-using-this-result-we-show-that-semantic-hierarchy-between-co-chunk-1.md
Created 2026-08-25T02:58:24+00:00
The results were empirically validated on Gemma-2B and LLaMA-3-8B using 593 noun and 364 verb WordNet synsets (retained if containing ≥50 words in the model vocabulary).
Summary
The underlying results weren't just theoretical; they were actually run on two real language models, Gemma-2B and LLaMA-3-8B, against a filtered set of word groupings pulled from WordNet (593 noun clusters and 364 verb clusters). The filtering kept only clusters where at least 50 words appeared in each model's vocabulary, so the tests exercised words the models can genuinely recognize rather than padding the evaluation with out-of-vocabulary noise.
Dependents
These beliefs depend on this one:
- IN multi-model-geometric-convergence — Both the polytope/orthogonality geometry (Park, validated on Gemma-2B and LLaMA-3-8B) and sparse feature structure (SAE, universal across architectures) converge on the finding that transformer representation spaces carry model-independent geometric invariants.