# What Happens During the Loss Plateau? Understanding Abrupt Learning in Transformers arXiv: https://arxiv.org/abs/2506.13688 ยท HTML: https://mlsanity.com/paper/2506.13688/ ## Key facts - Cited by 7 papers in the map (data through 14 Oct 2026). - Builds on 47 of its 59 references that are in the map. - First posted on arXiv 16 Jun 2025. ## Builds on (most cited first, 30 of 47) - [Training Verifiers to Solve Math Word Problems](https://mlsanity.com/paper/2110.14168/) (2021) - [Think you have Solved Question Answering? Try ARC, the AI2 Reasoning Challenge](https://mlsanity.com/paper/1803.05457/) (2018) - [The Curious Case of Neural Text Degeneration](https://mlsanity.com/paper/1904.09751/) (2019) - [Emergent Abilities of Large Language Models](https://mlsanity.com/paper/2206.07682/) (2022) - [HuggingFace's Transformers: State-of-the-art Natural Language Processing](https://mlsanity.com/paper/1910.03771/) (2019) - [Pythia: A Suite for Analyzing Large Language Models Across Training and Scaling](https://mlsanity.com/paper/2304.01373/) (2023) - [Progress measures for grokking via mechanistic interpretability](https://mlsanity.com/paper/2301.05217/) (2023) - [What Can Transformers Learn In-Context? A Case Study of Simple Function Classes](https://mlsanity.com/paper/2208.01066/) (2022) - [Grokking: Generalization Beyond Overfitting on Small Algorithmic Datasets](https://mlsanity.com/paper/2201.02177/) (2022) - [Towards Understanding Grokking: An Effective Theory of Representation Learning](https://mlsanity.com/paper/2205.10343/) (2022) - [Hidden Progress in Deep Learning: SGD Learns Parities Near the Computational Limit](https://mlsanity.com/paper/2207.08799/) (2022) - [SGD learning on neural networks: leap complexity and saddle-to-saddle dynamics](https://mlsanity.com/paper/2302.11055/) (2023) - [Signal Propagation in Transformers: Theoretical Perspectives and the Role of Rank Collapse](https://mlsanity.com/paper/2206.03126/) (2022) - [Sudden Drops in the Loss: Syntax Acquisition, Phase Transitions, and Simplicity Bias in MLMs](https://mlsanity.com/paper/2309.07311/) (2023) - [Teaching Arithmetic to Small Transformers](https://mlsanity.com/paper/2307.03381/) (2023) - [The mechanistic basis of data dependence and abrupt learning in an in-context classification task](https://mlsanity.com/paper/2312.03002/) (2023) - [The Evolution of Statistical Induction Heads: In-Context Learning Markov Chains](https://mlsanity.com/paper/2402.11004/) (2024) - [Explaining grokking through circuit efficiency](https://mlsanity.com/paper/2309.02390/) (2023) - [Grokking as the Transition from Lazy to Rich Training Dynamics](https://mlsanity.com/paper/2310.06110/) (2023) - [Learning to Break the Loop: Analyzing and Mitigating Repetitions for Neural Text Generation](https://mlsanity.com/paper/2206.02369/) (2022) - [Transformers need glasses! Information over-squashing in language tasks](https://mlsanity.com/paper/2406.04267/) (2024) - [What needs to go right for an induction head? A mechanistic study of in-context learning circuits and their formation](https://mlsanity.com/paper/2404.07129/) (2024) - [A Theoretical Analysis of the Repetition Problem in Text Generation](https://mlsanity.com/paper/2012.14660/) (2020) - [A Tale of Two Circuits: Grokking as Competition of Sparse and Dense Subnetworks](https://mlsanity.com/paper/2303.11873/) (2023) - [Dichotomy of Early and Late Phase Implicit Biases Can Provably Induce Grokking](https://mlsanity.com/paper/2311.18817/) (2023) - [Training Dynamics of Multi-Head Softmax Attention for In-Context Learning: Emergence, Convergence, and Optimality](https://mlsanity.com/paper/2402.19442/) (2024) - [Transformers learn through gradual rank increase](https://mlsanity.com/paper/2306.07042/) (2023) - [Grokking at the Edge of Numerical Stability](https://mlsanity.com/paper/2501.04697/) (2025) - [Training Dynamics of In-Context Learning in Linear Attention](https://mlsanity.com/paper/2501.16265/) (2025) - [Repetition In Repetition Out: Towards Understanding Neural Text Degeneration from the Data Perspective](https://mlsanity.com/paper/2310.10226/) (2023) ## Built on by (most cited first, 7 of 7) - [Arithmetic in the Wild: Llama uses Base-10 Addition to Reason About Cyclic Concepts](https://mlsanity.com/paper/2605.01148/) (2026) - [Convergent World Representations and Divergent Tasks](https://mlsanity.com/paper/2602.00533/) (2026) - [Understanding the Staged Dynamics of Transformers in Learning Latent Structure](https://mlsanity.com/paper/2511.19328/) (2025) - [Adaptive Engram Memory System for Indonesian Language Model: Generative AI Based on TOBA LM for Batak and Minang Language](https://mlsanity.com/paper/2603.10006/) (2026) - [Emergent Capabilities Arise Randomly from Learning Sparse Attention Patterns](https://mlsanity.com/paper/2606.25010/) (2026) - [From Syntax to Semantics: Unveiling the Emergence of Chirality in SMILES Translation Models](https://mlsanity.com/paper/2605.09949/) (2026) - [Training Dynamics of Transformers on Unbounded Inputs: Attention Learning Can Lag Output Layer Learning](https://mlsanity.com/paper/2610.20337/) (2026) ## Abstract (arXiv) Training Transformers on algorithmic tasks frequently demonstrates an intriguing abrupt learning phenomenon: an extended performance plateau followed by a sudden, sharp improvement. This work investigates the underlying mechanisms for such dynamics, primarily in shallow Transformers. We reveal that during the plateau, the model often develops an interpretable partial solution while simultaneously exhibiting a strong repetition bias in their outputs. This output degeneracy is accompanied by internal representation collapse, where hidden states across different tokens become nearly parallel. We further identify the slow learning of optimal attention maps as a key bottleneck. Hidden progress in attention configuration during the plateau precedes the eventual rapid convergence, and directly intervening on attention significantly alters plateau duration and the severity of repetition bias and representational collapse. We validate that these identified phenomena-repetition bias and representation collapse-are not artifacts of toy setups but also manifest in the early pre-training stage of large language models like Pythia and OLMo.