Pinned Loading

1. memory-bench memory-bench Public

   Controlled benchmark of memory mechanisms for transformers, built on nanochat

   Python

2. logprobe logprobe Public

   Detect normalization errors and entropy bias in LLM logprob data

   Rust

3. openai/parameter-golf openai/parameter-golf Public

   Train the smallest LM you can that fits in 16MB. Best model wins!

   Python 5k 3.4k

4. logprobe-calibration logprobe-calibration Public

   Calibration Under API Logprob Truncation — evidence from 16 model-provider conditions across OpenAI, Together AI, Groq, and self-hosted vLLM

   Python

5. smartbirds-research-site smartbirds-research-site Public

   Bird recognition from images and audio with YOLOv5s localization, ViT classification, BirdNET inference, and diagnostic views.

   TypeScript

6. FormalSLT FormalSLT Public

   Machine-verified statistical learning theory in Lean 4 — 45 modules, 19,521 lines, 0 sorry

   Lean 26 1