Shattered Compositionality: Counterintuitive Learning Dynamics of Transformers for Arithmetic
cs.LG, cs.AI
Submitted: 2026-01-30
Updated: 2026-11-10
Code: https://github.com/PeterXingyuZhao/Arithmetic_Transformer
Terminology
Sources
- LiveBench: A Challenging, Contamination-Limited LLM Benchmark
- Natural Emergent Misalignment from Reward Hacking in Production RL
- Grokking: Generalization Beyond Overfitting on Small Algorithmic Datasets
- Exploring LLM Reasoning Through Controlled Prompt Variations
- Understanding In-context Learning of Addition via Activation Subspaces
- A Theory for Emergence of Complex Skills in Language Models
- Task Generalization With AutoRegressive Compositional Structure: Can Learning From $D$ Tasks Generalize to $D^{T}$ Tasks?
- Learning Compositional Functions with Transformers from Easy-to-Hard Data
Related papers
- A Physics-Regulated Neural Framework for Learning 3D Grain Growth Dynamics
- FlexMoE: Learning One-for-All Multi-Budget Expert Pruning for Flexible MoE Deployment
- OptiMind: Improving Small Language Models for Optimization Modeling through Formulation Guidance and Multi-Turn Learning
- Staying in the Sweet Spot: Boost Reasoning Evolution via Responsive Rollout Difficulty Manipulation
- Examining the Impact of Data and Training Methods on Emergent and Subliminal Misalignment
- Rethinking Black-Box On-Policy Distillation: A Rubric-based Analysis