Pinned Loading

1. one-layer-priors one-layer-priors Public

   Do random-init priors predict which architecture solves a task? Applying Amid et al.'s Neural Network Prior Kernel as an architecture screen on the One Layer Deeper benchmark.

   Python

2. one-layer-deeper one-layer-deeper Public

   One Layer Deeper (Tilde Research x Core Automation): submissions + diagnostics. 100% on Easy and all ten Medium datasets with certified depth T=64; Hard root-caused.

   Python

3. prime-rl-environments prime-rl-environments Public

   Two deterministic RL environments for agent training—CTF-lite and git-surgery—with verifiable rewards, Prime Hub packaging, and baseline experiments.

   Python

4. website-cloning-agent website-cloning-agent Public

   An agentic system that turns any URL into a dev-ready multi-page Next.js clone. Playwright inspection at desktop + mobile, page-worker agents emit a Section IR, deterministic codegen writes a real …

   Python

5. reverse-llm reverse-llm Public

   A language model that reads forward English and generates character-reversed answers tail-first. Custom BPE on a mixed forward+reversed corpus, direction tokens for mode control; forked from nanochat.

   Python

6. autoft autoft Public

   Autonomous fine-tuning platform on Modal — research agent picks dataset, trains LoRA, streams live metrics

   TypeScript