AI Safety

My efforts in the area of AI SEA (Safety, Ethics and Alignment) have broadly been in these two areas:
Dataset auditing + bias mitigation and Model auditing and Red-Teaming + Adversarial attacks.

Here's a list of my recent publications spanning these two areas :

1. Model auditing, Red-Teaming and Adversarial attacks

   1. Beyond the imitation game: Quantifying and extrapolating the capabilities of language models 

   2. Auditing saliency cropping algorithms. Project page [Video]

   3. Model weight theft with just noise inputs: The curious case of the petulant attacker 

   4. Understanding adversarial robustness through loss landscape geometries 

   5. Vulnerability of deep learning-based gait biometric recognition to adversarial perturbations 

   6. On Lyapunov exponents and adversarial perturbation 

   7. On detecting adversarial inputs with entropy of saliency maps 

   8. On grey-box adversarial attacks and transfer learning 

   9. Art-attack! on style transfers with textures, label categories and adversarial examples 

   10. SIMUni: Sampling Impostors from Misfit Universal Background Models in accelerometric gait biometric verification 

   11.  Smile in the face of adversity much? A print based spoofing attack 

   12. OODles of ODDs:  The landscape of Out-of-distribution vulnerabilities of vision models

   13. Did They Direct the Violence or Admonish It? A Cautionary Tale on Contronomy, Androcentrism and Back-Translation Foibles [Video]

2. Dataset auditing and bias mitigation

   1. Multimodal datasets: misogyny, pornography, and malignant stereotypes 

   2. Large image datasets: A pyrrhic win for computer vision? 

   3. Into the laion’s den: Investigating hate in multimodal datasets 

   4. On Hate Scaling Laws For Data-Swamps

   5. The blood diamond effect in neural art: On ethically troublesome images of the imagenet dataset [Video]

   6. Covering up bias in CelebA-like datasets with Markov blankets: A post-hoc cure for attribute prior avoidance 

   7. JFT-300M: The mystery box