lm_understand-logs
lm_understand-logs is Complete logs of the experiments in the blog post 'LLMs can understand and write neural network weights' - RikiyaT/lm_understand-logs. It is ranked #149 on the AI Radar, in AI research, first seen 3 days ago and shared in 4 posts (54.8k views).
GitHub - RikiyaT/lm_understand-logs: Complete logs of the experiments in the blog post 'LLMs can understand and write neural network weights' · GitHub Skip to content Navigation Menu Sign in Search / Sign in Sign up Appearance settings You signed in with another tab or window. Reload to refresh your session. You signed out in another tab or window. Reload to refresh your session. You switched accounts on another tab or window. Reload to refresh your session. Dismiss alert {{ message }} Repository navigation lm_understand-logs Public Notifications You must be signed in to change notification settings Fork 0 ( 0 ) Star 13 ( 13 ) You must be…
What people said about lm_understand-logs on X
🤯天呐！GPT-6 Astra 居然直接“手改” 开源权重，Qwen3.5-0.8B小模型数学能力提升近6倍！无训练、无梯度！ MIT CSAIL 团队让 GPT-6 Astra 直接读取并编辑 Qwen3.5-0.8B 的权重（不是微调！），仅改动 0.016% 的参数，就解决了官方模型卡都警告的推理死循环问题，数学能力最高提升约 6倍，指令跟随也涨了约 3倍！ 这就是他们提出的 Weight-Space Programming（权重空间编程）：前沿大模型已经能原生读写神经网络权重了！ 核心亮点 🔹只改了 0.016% 权重（约13万个参数），无梯度、无训练 🔹修复 Qwen 官方已知的推理循环问题…
— @NFT_Chen, 3 days ago · 193 likes · see the post
pretty crazy to me that this kind of behavior just emerge without any specific training
— @eliebakouch, 3 days ago · 141 likes · see the post
the advantage of open models is that you can read what they think Kimi K3.1 Pro contemplates the network's raw weights to determine the game the policy was trained on don't show this to Yud…
— @teortaxesTex, 3 days ago · 155 likes · see the post
Alternatives to lm_understand-logs
- alphaXiv — Now on @askalphaxiv
- flashprove — Further progress in provable inference will likely be an important part of improving trust in AI deployment. To that…
- moe-blogpost — A hands-on guide to training Mixture-of-Experts models
- liberai.com — Technical blog
- Science's Last Exam — Can AI do research, end to end? Science's Last Exam defines and measures whether an AI agent can carry a research goal…
- science-last-exam — Science's Last Exam. Contribute to Gen-Verse/Science-Last-Exam development by creating an account on GitHub.
lm_understand-logs in numbers
- Rank on the AI Radar: #149 of 10395
- Shared in 4 posts by 3 accounts: @NFT_Chen, @eliebakouch, @teortaxesTex
- 54.8k views on those posts
- First seen 3 days ago, last shared 3 days ago
- Pricing seen by AGK Intelligence: open source
- Market: AI research
FAQ
What is lm_understand-logs?
Complete logs of the experiments in the blog post 'LLMs can understand and write neural network weights' - RikiyaT/lm_understand-logs It was first shared on X 3 days ago and is ranked #149 on the AI Radar.
Is lm_understand-logs free?
It is open source.
Who shared lm_understand-logs?
3 accounts on X, including @NFT_Chen, @eliebakouch, @teortaxesTex, in 4 posts totalling 54.8k views.
Every post is read and classified by AGK Intelligence: what it is, which market it belongs to, and whether the link is a real tool. 65.7k posts from 5k X accounts over the last 14 days, 10.4k tools, 19 markets. New posts are collected several times an hour and the full ranking is recomputed every hour — last update 2026-11-06 14:10 UTC. Full method.