Same Function, Different Mechanism: Implementation-Local Geometry Beyond Accuracy
cs.LG
Submitted: 2026-10-31
Updated: 2026-10-31
Terminology
Sources
- Geometric Deep Learning: Grids, Groups, Graphs, Geodesics, and Gauges
- Geometry of Program Synthesis
- Causal Abstraction: A Theoretical Foundation for Mechanistic Interpretability
- Loss Landscape Degeneracy and Stagewise Development in Transformers
- Tracr: Compiled Transformers as a Laboratory for Interpretability
- MIB: A Mechanistic Interpretability Benchmark
- Programs as Singularities
- Progress measures for grokking via mechanistic interpretability
- Grokking: Generalization Beyond Overfitting on Small Algorithmic Datasets
- Differentiation and Specialization of Attention Heads via the Refined Local Learning Coefficient
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