The Residual Stream's Effective Depth
cs.LG
Submitted: 2026-09-25
Updated: 2026-09-25
Terminology
Sources
- Eliciting Latent Predictions from Transformers with the Tuned Lens
- Do Language Models Use Their Depth Efficiently?
- The Llama 3 Herd of Models
- LayerSkip: Enabling Early Exit Inference and Self-Speculative Decoding
- Gemma: Open Models Based on Gemini Research and Technology
- OLMo: Accelerating the Science of Language Models
- The Unreasonable Ineffectiveness of the Deeper Layers
- Mistral 7B
- Shortened LLaMA: Depth Pruning for Large Language Models with Comparison of Retraining Methods
- The Remarkable Robustness of LLMs: Stages of Inference?
- ShortGPT: Layers in Large Language Models are More Redundant Than You Expect
- The FineWeb Datasets: Decanting the Web for the Finest Text Data at Scale
- Mixture-of-Depths: Dynamically allocating compute in transformer-based language models
- Does Representation Matter? Exploring Intermediate Layers in Large Language Models
- The Curse of Depth in Large Language Models
Related papers
- Polynomial-Augmented Neural Networks (PANNs) with Weak Orthogonality Constraints for Enhanced Function and PDE Approximation
- AIRL-S: Unifying Reinforcement Learning and Search-Based Test-Time Scaling via Adversarial Inverse Reinforcement Learning
- Transformers as Bayesian In-Context Experimenters: Smoothness-Adaptive Efficient ATE Estimation
- Convergence issues in Relational Concept Analysis based on AOC-posets
- Beliefs Beyond Posteriors: Local-Consistency Optimisation for Bayesian Neural Networks
- Understanding Diffusion Models via Ratio-Based Function Approximation with SignReLU Networks