stefdesabbata/geospatial-mechanistic-interpretability
Geospatial Mechanistic Interpretability of Large Language Models
This project helps researchers and academics understand how large language models (LLMs) process geographical information internally. It takes LLM responses to placenames and applies spatial analysis techniques to reveal how these models represent and "think" about locations. The output provides insights into the internal mechanisms of LLMs concerning geospatial data, useful for academics studying AI behavior.
No commits in the last 6 months.
Use this if you are a researcher or academic specifically interested in dissecting the internal workings of large language models when they deal with geographic data, aiming to understand their spatial reasoning.
Not ideal if you are looking for a tool to directly improve an LLM's geographical accuracy or to simply apply LLMs to geospatial tasks without needing to understand their internal representations.
Stars
18
Forks
—
Language
Jupyter Notebook
License
MIT
Category
Last pushed
May 12, 2025
Commits (30d)
0
Get this data via API
curl "https://pt-edge.onrender.com/api/v1/quality/llm-tools/stefdesabbata/geospatial-mechanistic-interpretability"
Open to everyone — 100 requests/day, no key needed. Get a free key for 1,000/day.
Higher-rated alternatives
filipnaudot/llmSHAP
llmSHAP: a multi-threaded explainability framework using Shapley values for LLM-based outputs.
microsoft/automated-brain-explanations
Generating and validating natural-language explanations for the brain.
CAS-SIAT-XinHai/CPsyCoun
[ACL 2024] CPsyCoun: A Report-based Multi-turn Dialogue Reconstruction and Evaluation Framework...
wesg52/universal-neurons
Universal Neurons in GPT2 Language Models
ICTMCG/LLM-for-misinformation-research
Paper list of misinformation research using (multi-modal) large language models, i.e., (M)LLMs.