A single middle transformer layer trained in isolation recovers most RL post-training gains in LLMs, with gains concentrated in middle layers across models, algorithms, and tasks.
arXiv preprint arXiv:2510.02091 , year=
5 Pith papers cite this work. Polarity classification is still indexing.
5
Pith papers citing it
years
2026 5representative citing papers
Internal attention in LLMs shows a bell-curve relevance distribution across layers, enabling Selective-ICR that cuts inference latency 30-50% and lets an 8B zero-shot model match 14B RL re-rankers on BRIGHT.
Task-aware pruning improves OOD model performance by realigning distorted OOD layerwise norm and pairwise-distance profiles with the task-adapted geometry observed on ID inputs.
Internal LLM artifacts can be used to build classifiers that identify incorrect predictions on legal classification tasks.
citing papers explorer
-
Is One Layer Enough? Training A Single Transformer Layer Can Match Full-Parameter RL Training
A single middle transformer layer trained in isolation recovers most RL post-training gains in LLMs, with gains concentrated in middle layers across models, algorithms, and tasks.
-
Where Relevance Emerges: A Layer-Wise Study of Internal Attention for Zero-Shot Re-Ranking
Internal attention in LLMs shows a bell-curve relevance distribution across layers, enabling Selective-ICR that cuts inference latency 30-50% and lets an 8B zero-shot model match 14B RL re-rankers on BRIGHT.
-
TAPIOCA: Why Task- Aware Pruning Improves OOD model Capability
Task-aware pruning improves OOD model performance by realigning distorted OOD layerwise norm and pairwise-distance profiles with the task-adapted geometry observed on ID inputs.
-
Peeking Inside LLMs: Leveraging Internal Artifacts of LLMs for Enhancing Reliability in Legal Classification
Internal LLM artifacts can be used to build classifiers that identify incorrect predictions on legal classification tasks.
- On the Limits of Layer Pruning for Generative Reasoning in Large Language Models