I'm Matt Wood, and this is For Your Information. A live list of riffs and links for you and your agent, drawn from what I'm reading, noticing, questioning, concluding, and revising.
Evaluating LLM Judge Agreement and ReliabilityBoth deal with classification reliability and evaluation - Numberwang NN satirizes what 'correct' classification even means when the ground truth is arbitrary, mirroring real concerns about LLM judge agreement
The Emergent Symbolic Structure of Artificial Neural NetworksBoth explore neural networks applied to abstract/symbolic classification tasks - Numberwang NN classifies arbitrary numbers while the other examines emergent symbolic structures in ANNs
Bespoke: A Programming Language for People Who Say PleaseBoth are whimsical, comedy-derived technical projects that use real engineering (neural nets / programming language design) to implement absurdist concepts from British humor
Comparing 11 Different AI ModelsBoth involve benchmarking/evaluating models on classification tasks, though Numberwang deliberately subverts the notion that benchmark targets are meaningful
Challenges
The End of MathematicsNumberwang deliberately parodies mathematical meaninglessness - a neural network trained to classify 'Numberwang' satirically challenges the notion that mathematics has inherent structure or that ML classification tasks are always meaningful
mattwood.fyi