Context Staircase: Signature-Aligned Dynamics of Token Embeddings under Small Initialization
cs.LG, cs.CL
Submitted: 2026-08-31
Updated: 2026-08-31
Terminology
Sources
- Grokking modular arithmetic
- Efficient Estimation of Word Representations in Vector Space
- GPT-4 Technical Report
- Megatron-LM: Training Multi-Billion Parameter Language Models Using Model Parallelism
- LLaMA: Open and Efficient Foundation Language Models
- Order Matters in Hallucination: Reasoning Order as Benchmark and Reflexive Prompting for Large-Language-Models
- An overview of condensation phenomenon in deep learning
- Number Cookbook: Number Understanding of Language Models and How to Improve It
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: Local-Consistency Optimisation for Bayesian Neural Networks
- Understanding Diffusion Models via Ratio-Based Function Approximation with SignReLU Networks