If you train a model for too long, it may overfit it's training data. Not surprising, this has been know for like forever. But did you know you can detect the signatures of overfitting in the layer weight matrices directly, without needing access to any data (train or test) ? In our recent paper (with hari kishan prakash ), 𝐋𝐚𝐭𝐞-𝐒𝐭𝐚𝐠𝐞 𝐆𝐞𝐧𝐞𝐫𝐚𝐥𝐢𝐳𝐚𝐭𝐢𝐨𝐧 𝐂𝐨𝐥𝐥𝐚𝐩𝐬𝐞 𝐢𝐧 𝐆𝐫𝐨𝐤𝐤𝐢𝐧𝐠: 𝐃𝐞𝐭𝐞𝐜𝐭𝐢𝐧𝐠 𝐀𝐧𝐭𝐢-𝐆𝐫𝐨𝐤𝐤𝐢𝐧𝐠 𝐰𝐢𝐭𝐡 𝐖𝐞𝐢𝐠𝐡𝐭𝐖𝐚𝐭𝐜𝐡𝐞𝐫, we show this explicitly in 2 different classic grokking experiments. And the overfitting we see is very different from what has been seen before! 📄 paper: arxiv.org/abs/2602.02859 📈 𝐖𝐡𝐚𝐭 𝐭𝐡𝐢𝐬 𝐩𝐥𝐨𝐭 𝐬𝐡𝐨𝐰𝐬: Grokking → Stability → Anti-Grokking This figure below tracks training accuracy (red), test accuracy (purple), and WeightWatcher correlation traps (blue) while training for very long times. 𝐏𝐡𝐚𝐬𝐞 𝟏 — Memorization (pre-grokking) Training accuracy rises rapidly while test accuracy remains low. The model is fitting the training data without extracting the underlying structure. Correlation traps are minimal and largely uninformative. 𝐏𝐡𝐚𝐬𝐞 𝟐 — Grokking Test accuracy suddenly jumps to match training accuracy. The model transitions from memorization to true generalization. Correlation traps remain near zero, indicating stable, well-conditioned internal representations. 𝐏𝐡𝐚𝐬𝐞 𝟑 — Late-stage instability (anti-grokking) Despite perfect training accuracy, test accuracy degrades over time. At the same time, correlation traps increase sharply and spread. Generalization collapses after it was achieved. The model is overfit 𝐊𝐞𝐲 𝐭𝐚𝐤𝐞𝐚𝐰𝐚𝐲 So what ? It turns out, a lot of open-source LLMs, like OpenAI's GPT OSS 20B and 120B , show the exact same signatures! And a ton of them If you are training or fine-tuning your own models, watch out! You might be overfitting your data, even if you are following current NN best practices. Want to learn more? Check out the WeightWatcher project 🌐 weightwatcher.ai 𝐖𝐞𝐢𝐠𝐡𝐭𝐖𝐚𝐭𝐜𝐡𝐞𝐫 𝐢𝐬 𝐚 𝐨𝐧𝐞-𝐨𝐟-𝐚-𝐤𝐢𝐧𝐝 𝐦𝐮𝐬𝐭-𝐡𝐚𝐯𝐞 𝐭𝐨𝐨𝐥 𝐟𝐨𝐫 𝐚𝐧𝐲𝐨𝐧𝐞 𝐭𝐫𝐚𝐢𝐧𝐢𝐧𝐠, 𝐝𝐞𝐩𝐥𝐨𝐲𝐢𝐧𝐠, 𝐨𝐫 𝐦𝐨𝐧𝐢𝐭𝐨𝐫𝐢𝐧𝐠 𝐃𝐞𝐞𝐩 𝐍𝐞𝐮𝐫𝐚𝐥 𝐍𝐞𝐭𝐰𝐨𝐫𝐤𝐬 (𝐃𝐍𝐍𝐬). And you need help with AI, reach out. hashtag#TalkToChuck P.S. I'll be giving a talk at USF right here in SF (over by the GG Park) in 2 weeks on the weightwatcher project. Hope to see you there!