arXiv (All)
open access
Transformers Learn Two Competing Sorting Algorithms, and Only One Length-Generalizes
We show that two-layer transformers trained to near-zero loss on a sorting task converge to one of two mechanistically distinct solution classes depending on their initialization. One solution class consists of solutions that do not use the second attention layer to sort, but also fail to length-gen
Conceptio Open Archive
Analyze & explore related research in Conceptio
Connect this record directly into your research workflow or AI agent. Explore citations across open repositories, query full text, and export structured proof bundles.
This document is indexed with metadata only: full text is not available in the archive for this record. Open the official source ↗
Related documents
- Official Review · ICLR 2025#787156
- Counting with Attention and LayerNorm: Norm Scaling and Representation Geometry#1341542
- A Unified Perspective on the Dynamics of Deep Transformers#1120649
- Compressing History into Memory: Distilling Transformers into Recurrent Transformers#1156396
- Attention Sinks Induce Gradient Sinks: Massive Activations as Gradient Regulators in Transformers#1400865
Record · ID 1402362
Retrieved via Conceptio: every document is proof-bundled with source, license, and retrieval metadata.
Retrieved via Conceptio: every document is proof-bundled with source, license, and retrieval metadata.
Bring this data into your agent: Connect via MCP or API · 20 free credits to start · no card required
Read the Docs → · View CLI source on GitHub · Conceptio Open Knowledge Archive
Read the Docs → · View CLI source on GitHub · Conceptio Open Knowledge Archive