GTA (Guess The Algorithm) Benchmark - A tool for testing AI reasoning capabilities
Modular Arithmetic Challenge. Neural induction of exact (a x b) mod p through abacus embeddings, algorithmic scratchpads and grokking, for the SAIR Foundation competition.
A sketch-guided multi-objective alignment framework for improving the correctness, efficiency, and algorithmic reasoning of code language models.
An evaluation of (multitask) algorithmic reasoning
Reproducible experiments on executable-semantics acquisition in language models, including training-order-induced output collapse and memorization without held-out modulo generalization.
Is the graph doing anything? A sourced survey of recurrent self-routing neural graphs as an alternative to layered networks (SR-2026-006)
Multi-agent collaborative reasoning system using Microsoft AutoGen to solve and benchmark complex Data Structures & Algorithms.
Neural cellular automata for Go liberty counting: stable rollouts, size-generalization failures, matched baselines, and reproducible evidence.
Fourier-analysis replications and extensions for grokking and algorithmic reasoning
Algorithmic reasoning and performance engineering for AI coding agents. Discover better algorithms, optimize complexity, and verify performance improvements.
Challenge your AI's algorithmic thinking with GTA Benchmark! Reverse-engineer transformations from input-output pairs. Join now! 🐙🚀
Reverse-engineering a custom mini-transformer to mathematically prove how it learns and computes arithmetic.
A three-seed study of grokking in modular division: delayed generalization, late training instability, and reproducible PyTorch analysis.
A TensorFlow CNN project analyzing the limitations of neural networks on algorithmic reasoning tasks using binary matrix datasets.
Modern PyTorch reproduction and graph extension of Neural Programmer-Interpreters
Analyze code snippets with an AI assistant to find design flaws, suggest refactoring, generate tests, and audit SOLID principles efficiently.
To associate your repository with the algorithmic-reasoning topic, visit your repo's landing page and select "manage topics."