Forking: Sudden Overfitting Under Replay

cs.LG

Submitted: 2026-09-30

Updated: 2026-09-30

Comments: 42 pages, 22 figures. Code and reproduction materials: https://github.com/guoshaoyang-pku/forking/tree/release

Code: https://github.com/guoshaoyang-pku/forking

License: http://arxiv.org/licenses/nonexclusive-distrib/1.0/

Sources

• Memory Grafting: Scaling Language Model Pre-training via Offline Conditional Memory
• DeepSeek-V4.1-Flash: Pushing the Limits of KV Cache Compression
• Notes on the occupancy problem with infinitely many boxes: general asymptotics and power laws
• Scaling Laws and Interpretability of Learning from Repeated Data
• Data Scarcity and Model Sparsity: Mixtures-of-Experts Overfit More to Repeated Data
• The AI Scientist: Towards Fully Automated Open-Ended Scientific Discovery
• N-Grammer: Augmenting Transformers with latent n-grams
• The AI Scientist-v2: Workshop-Level Automated Scientific Discovery via Agentic Tree Search

Related papers

• Polynomial-Augmented Neural Networks (PANNs) with Weak Orthogonality Constraints for Enhanced Function and PDE Approximation
• AIRL-S: Unifying Reinforcement Learning and Search-Based Test-Time Scaling via Adversarial Inverse Reinforcement Learning
• Transformers as Bayesian In-Context Experimenters: Smoothness-Adaptive Efficient ATE Estimation
• Convergence issues in Relational Concept Analysis based on AOC-posets
• Beliefs Beyond Posteriors: The Predictive Cost of ELBO Binding
• Understanding Diffusion Models via Ratio-Based Function Approximation with SignReLU Networks