<html><head><meta http-equiv="content-type" content="text/html; charset=utf-8"></head><body style="overflow-wrap: break-word; -webkit-nbsp-mode: space; line-break: after-white-space;"><meta http-equiv="content-type" content="text/html; charset=utf-8"><div style="overflow-wrap: break-word; -webkit-nbsp-mode: space; line-break: after-white-space;"><div><span style="caret-color: rgb(0, 0, 0); color: rgb(0, 0, 0);">(Apologies if you receive multiple copies of this message)</span></div><div><br></div>Our next 1-hour webinar will take place <b>Wednesday, June 4, 2025 at 3:00 PM CEST </b>(Europe/Paris) with Dr. Deepti Ghadiyaram (Boston University, US).<div><br>RSVP here to join: <a href="https://cassyni.com/events/KrZokXADhoLa388YBkWx4V?cb=0.blj3">https://cassyni.com/events/KrZokXADhoLa388YBkWx4V?cb=0.blj3</a> <br><br></div><div><b><span>Title: Revealing and Leveraging the Visual Information in Diffusion Models</span><span><br></span></b><span><br></span></div><div><span><b>Abstract</b>: Generating high-quality photo-realistic and creative visual content using diffusion models is a thriving area of research. In this talk, I will focus not on the generation process, but on understanding and leveraging the rich visual semantic information represented within diffusion models. Specifically, I will present our work that uses mechanistic interpretability tools such as k-sparse autoencoders (k-SAE) to probe various layers and denoising timesteps of different diffusion architectures. Next, I will present how to uncover monosemantic interpretable concepts pertaining to safety and photographic styles and steer the generation process thereby offering more controllability to users.</span><span><br></span><span><br></span></div><div><span><b>Bio</b>: Deepti is an Assistant Professor at Boston University in the Department of Computer Science and also a Member of Technical Staff at Runway. Her research interests are on topics pertaining to safe and interpretable computer vision, improving realism in generative video models, and human actions. Prior to joining Boston University, she was a Senior Research Scientist at Fundamental AI Research (FAIR) in Meta AI. She has served as a program chair for NeurIPS 2022 Dataset and Benchmarks track, hosted several tutorials and organized workshops and an area chair for several years at CVPR, ICCV, ECCV, ACCV, AAAI, and NeurIPS.</span><span><br></span><span><br></span></div><div><span>We look forward to your attendance.</span></div><div><span><br></span><span><div>
—<br>__________________________________________<br>Dr. Giuseppe Valenzise<br>CNRS Researcher<br>Laboratoire des Signaux et Systèmes (L2S, UMR 8506)<br>CNRS - CentraleSupelec - Université Paris-Saclay<br>3, rue Joliot Curie<br>91192 Gif-sur-Yvette Cedex, France<br>https://l2s.centralesupelec.fr/u/valenzise-giuseppe/<br><br>General Chair - ICME 2025<br><br>Chair of the IEEE SPS Multimedia Signal Processing Technical Committee (MMSP TC)<br><br>Editor in Chief EURASIP Journal on Image and Video Processing<br><span></span><span></span><span></span><span></span><span></span><span></span><br class="Apple-interchange-newline"><span></span><span></span><span></span><span></span><span><img alt="signature_banner.png" src="cid:BDFBF4EB-9312-4804-9085-0D76C8FA74F5"></span><br class="Apple-interchange-newline"><br class="Apple-interchange-newline"><br class="Apple-interchange-newline"><br class="Apple-interchange-newline">~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~<br>“Immersive Video Technologies”<br>https://www.elsevier.com/books/immersive-video-technologies/valenzise/978-0-323-91755-1<br><br>
</div></span><span>
</span><span><br></span><div></div></div></div></body></html>