Publications

2026

1. ICLR

   Detecting and Mitigating Memorization in Diffusion Models through Anisotropy of the Log-Probability

   Rohan Asthana and Vasileios Belagiannis

   In International Conference on Learning Representations, 2026

   Abs HTML Code Website

   We propose an anisotropy-based metric for detecting and mitigating memorization in diffusion models (e.g., Stable Diffusion). Our method is 5-8x faster than prior approaches and requires only 2 forward passes.

2025

1. TMLR

   Dextr: Zero-Shot Neural Architecture Search with Singular Value Decomposition and Extrinsic Curvature

   Rohan Asthana, Joschua Conrad, Maurits Ortmanns, and 1 more author

   Transactions on Machine Learning Research, Aug 2025

   Abs HTML Code

   We introduce Dextr, a zero-shot NAS proxy that uses singular value decomposition and extrinsic curvature to rank neural architectures without any labeled data, requiring only a single unlabeled sample.

2. IEEE

   PSumSim: A Simulator for Partial-Sum Quantization in Analog Matrix-Vector Multipliers

   Joschua Conrad, Simon Wilhelmstätter, Holger Mandry, and 5 more authors

   In 2025 IEEE International Symposium on Circuits and Systems (ISCAS), 2025

   HTML

2024

1. TMLR

   Multi-conditioned Graph Diffusion for Neural Architecture Search

   Rohan Asthana, Joschua Conrad, Youssef Dawoud, and 2 more authors

   Transactions on Machine Learning Research, Mar 2024

   Abs HTML Code

   We propose DiNAS, a conditional graph diffusion model capable of generating valid neural architectures in under 0.2 seconds, evaluated across 6 NAS benchmarks.

2. IEEE

   Differentiable Cost Model for Neural-Network Accelerator Regarding Memory Hierarchy

   Joschua Conrad, Simon Wilhelmstätter, Rohan Asthana, and 2 more authors

   IEEE Transactions on Circuits and Systems I: Regular Papers, 2024

   HTML
3. IEEE

   Too-Hot-to-Handle: Insights into Temperature and Noise Hyperparameters for Differentiable Neural-Architecture-Searches

   Joschua Conrad, Simon Wilhelmstätter, Rohan Asthana, and 2 more authors

   In 2024 IEEE 6th International Conference on AI Circuits and Systems (AICAS), 2024

   HTML
4. IEEE

   Confidence Estimation and Boosting for Dynamic-Comparator Transient-Noise Analysis

   Joschua Conrad, John G. Kauffman, Simon Wilhelmstätter, and 3 more authors

   In 2024 22nd IEEE Interregional NEWCAS Conference, 2024

   HTML

2021

1. ETT

   Fake News Detection using Deep Learning Models: A Novel Approach

   Sachin Kumar, Rohan Asthana, S. Upadhyay, and 2 more authors

   Transactions on Emerging Telecommunications Technologies, 2021

   HTML