#interpretability

• 22 Oct 2026 · Paper summary

  Paper summary: Spectral-Spatial Interpretation

  Spectral-Spatial Interpretation explains multispectral image models with a full position × channel attribution matrix, using a local Shapley game at each position, so it shows which channels matter where.

• 11 Mar 2025 · Paper summary

  Paper summary: Feature-Level Insights into Artificial Text Detection with Sparse Autoencoders

  Using sparse-autoencoder features from Gemma-2-2B's residual stream to detect and explain AI-generated text, revealing a distinct LLM writing style, especially in information-dense domains.

• 24 Feb 2025 · Paper summary

  Paper summary: LLM-Microscope: Uncovering the Hidden Role of Punctuation in Context Memory of Transformers

  Tokens we treat as filler (punctuation, articles, stop words) carry a surprising share of an LLM's contextual memory; removing them hurts MMLU and long-context scores, and the paper releases a toolkit to measure this.

• 27 Mar 2024 · Paper summary

  Paper summary: The Unreasonable Ineffectiveness of the Deeper Layers

  Removing up to about half of the layers of some open LLMs, chosen by layer similarity and followed by a little QLoRA "healing", barely hurts question-answering performance.