# What Layers When: Learning to Skip Compute in LLMs with Residual Gates

arXiv: https://arxiv.org/abs/2510.13876 · HTML: https://mlsanity.com/paper/2510.13876/

## Key facts

- Cited by 6 papers in the map (data through 14 Oct 2026).
- Placed in Early exit and edge inference, one of 12 directions in Efficient LLM inference and serving.
- First posted on arXiv 13 Oct 2025.

## Placed in

- Efficient LLM inference and serving / [Early exit and edge inference](https://mlsanity.com/inference/early-exit-and-edge-inference/)

## Built on by (most cited first, 6 of 6)

- [Locate, Steer, and Improve: A Practical Survey of Actionable Mechanistic Interpretability in Large Language Models](https://mlsanity.com/paper/2601.14004/) (2026)
- [Towards Understanding Layer Contributions in Tabular In-Context Learning Models](https://mlsanity.com/paper/2511.15432/) (2025)
- [ProxyKV: Cross-Model Proxy Pruning for Efficient Long-Context LLM Inference](https://mlsanity.com/paper/2605.16360/) (2026)
- [Reachability Is Not Realization: Tracing the Sources of LLM Benchmark Gains](https://mlsanity.com/paper/2608.03219/) (2026)
- [Programs-of-Layers in LLMs through the Lens of Cortical Areas](https://mlsanity.com/paper/2609.31360/) (2026)
- [Do Dynamic Routers Need Memory? HeRo: History-Aware Routing for Efficient LLM Inference](https://mlsanity.com/paper/2609.08189/) (2026)

## Abstract (arXiv)

We introduce GateSkip, a simple residual-stream gating mechanism that enables token-wise layer skipping in decoder-only LMs. Each Attention/MLP branch is equipped with a sigmoid-linear gate that condenses the branch's output before it re-enters the residual stream. During inference we rank tokens by the gate values and skip low-importance ones using a per-layer budget. While early-exit or router-based Mixture-of-Depths models are known to be unstable and need extensive retraining, our smooth, differentiable gates fine-tune stably on top of pretrained models. On long-form reasoning, we save up to 15% compute while retaining over 90% of baseline accuracy. For increasingly larger models, this tradeoff improves drastically. On instruction-tuned models we see accuracy gains at full compute and match baseline quality near 50% savings. The learned gates give insight into transformer information flow (e.g., BOS tokens act as anchors), and the method combines easily with quantization, pruning, and self-speculative decoding.