←
Understandable Machine Intelligence Lab186 posts
Understandable Machine Intelligence Lab avatar

Understandable Machine Intelligence Lab

@UMI_Lab_AI

Understandable Machine Intelligence Lab: We bring #explainable #AI to the next level. Part of @LeibnizATB, Ex @TUBerlin, funded by @BMBF_Bund #XAI

◷ Joined November 2020▧ 59 Photos and videos
Understandable Machine Intelligence Lab avatar
Understandable Machine Intelligence Lab@UMI_Lab_AI·22 Sep 2025

MRR, NeurIPS Workshop (2024)  ⚠️ The Price of Freedom: An Adversarial Attack on Interpretability Evaluation. Shows how parameterizarion can be exploited to manipulate evals, and proposes ranking-based mitigation.  📚 openreview.net/pdf?id=BpJK5l… 👉 github.com/Wickstrom/quantit…

Understandable Machine Intelligence Lab avatar
Understandable Machine Intelligence Lab@UMI_Lab_AI·22 Sep 2025

NoiseGrad, AAAI (2022) ✨ NoiseGrad — Enhancing explanations by adding stochasticity in the weight space. Improving the faithfulness and robustness of local + global explanations, extending the SmoothGrad method. 📚 ojs.aaai.org/index.php/AAAI/… 👉 github.com/understandable-ma…

Understandable Machine Intelligence Lab avatar
Understandable Machine Intelligence Lab@UMI_Lab_AI·22 Sep 2025

Quantus, JMLR (2023) ⚙️ Quantus — an open-source library with 35+ metrics for reproducible evaluation of explanations. Now with 60k+ downloads & 600+ ⭐️, it set the foundation for systematic eval! 📚 jmlr.org/papers/v24/22-0142.… 👉 github.com/understandable-ma…

Understandable Machine Intelligence Lab avatar
Understandable Machine Intelligence Lab@UMI_Lab_AI·22 Sep 2025

🔊 Not to miss …. last month @anna_hedstroem defended her PhD “Evaluation-centric advances in neural model interpretability” at TU Berlin — with distinction! ✨🧠💻☕️ Here’s a thread of a selection of Anna’s evaluation-centric interpretability work + what comes next. 🧵

Post attachment
Understandable Machine Intelligence Lab avatar
Understandable Machine Intelligence Lab@UMI_Lab_AI·24 Jul 2025

🧐 DORA: Exploring Outlier Representations in Deep Neural Networks (TMLR 2023)
A framework for analyzing & detecting learned representations in neural networks. 👉 arxiv.org/abs/2206.04530