[ACL 2025] 💻🔍 How Programming Concepts and Neurons Are Shared in Code Language Models
-
Updated
Mar 7, 2025 - Jupyter Notebook
#
logitlens
Here are 4 public repositories matching this topic...
Projektarbeit über Visualisierung der Token-Wichtigkeit zur Aufdeckung und Analyse von Prompt Injections in Sprachmodelle
-
Updated
Nov 29, 2025 - Jupyter Notebook
Mechanistic analysis of a GPT-2–like model exploring the compositionality gap in transformers. Using Logit Lens and Causal Tracing, the study identifies and overcomes a deep-layer bottleneck via dataset enhancement addressing the stated Compositionality Gap (NeurIPS24).
tokenizer hpc-clusters explainable-ai ai-research explainability compositionality neurips llm logitlens mechanistic-interpretability llm-training causal-tracing gpt-2-model ai-deployment neurips-paper
-
Updated
Aug 9, 2026 - Python
This Alignment Jam Hackathon project explores whether the concept of "logit lens" applies to the encoder and decoder layers in Whisper, an end-to-end speech recognition model.
-
Updated
Jan 3, 2023 - Jupyter Notebook
Add this topic to your repo
To associate your repository with the logitlens topic, visit your repo's landing page and select "manage topics."