Searching for an Authentic Signature Dataset with Less than 4-bit Biased Nonces for Learning

Hey everyone! I'm diving into how nonce bias might expose ECDSA private keys. I’m really keen on finding actual datasets where the nonces in Bitcoin or similar systems are biased by under 4 bits like 1-bit, 2-bit, or 3-bit biases. Here's what I'm after: - Real signatures where the nonce k has a bias - Known private key (d) for practicing recovery - Clearly labeled bias type (MSB or LSB) - Dataset formatted as simple JSON (like the example below) If anyone has played around with or created datasets like this, especially where a 1-bit bias led to recovering the private key, I'd be super interested in hearing about it. I’m not looking for anyone's private or wallet keys, just some clean test vectors for educational purposes. I've also come across this link: https://github.com/bitlogik/lattice-attack I'm curious about how low-bit biases (like 1-3 bits) can expose d, how many signatures you really need per bias level, what attack methods (like HNP, lattice, or ML) are effective in practice, and how to recreate or verify the pipeline of bias leading to d leak. Thanks a ton to anyone who’s willing to share or chat about this! I’m only doing this for research and learning and would gladly share my own findings and scripts if there's interest. Also, if you've ever worked on recovering private keys from biased nonces, I want to know: ✅ How many biased signatures did you need to get d for: 1-bit bias? 2-bit bias? 3-bit bias? 🧠 What method did you use for the attack? Hidden Number Problem (HNP)? Lattice methods (LLL, BKZ)? Custom optimizations? ⚙️ What hardware did you use? CPU (how many cores/threads)?

i need 1-bit, data.json if successful recover d some one provide dm   Where can find 1-bit data.json available publicly the data was needed to solve d.