ResearchTrend Connect | FedML · VLM · LLMAG
Details
### Join us for a dynamic 3-hour online event featuring three back-to-back ResearchTrend Connect sessions, each spotlighting cutting-edge developments in a key area of machine learning and AI research: Federated Machine Learning (FedML), Vision-Language Models (VLM), and Large-Language Model Agents (LLMAG).
### Agenda
ResearchTrend Connect | FedML:
16:00 - 16:05 (UTC) Introduction
16:05 - 16:15 (UTC): Ertuğrul Keçeci-A State Alignment-Centric Approach to Federated System Identification: The FedAlign Framework
16:20 - 16:30 (UTC): Youngjoon Leet-Revisit the Stability of Vanilla Federated Learning Under Diverse Conditions
16:35 - 17:00 (UTC): Discussion
ResearchTrend Connect | VLM:
17:00 - 17:05 (UTC) Introduction
17:05 - 17:15 (UTC): Yue Zhao-QLIP: Text-Aligned Visual Tokenization Unifies Auto-Regressive Multimodal Understanding and Generation
17:20 - 17:30 (UTC): Dominik Schnaus-It's a (Blind) Match! Towards Vision-Language Correspondence without Parallel Data
17:35 - 18:00 (UTC): Discussion
ResearchTrend Connect | LLMAG:
18:00 - 18:05 (UTC) Introduction
18:05 - 18:15 (UTC): Mert Cemri-Why Do Multi-Agent LLM Systems Fail?
18:20 - 18:30 (UTC): Rana Muhammad Shahroz Khan-$\textit{Agents Under Siege}$: Breaking Pragmatic Multi-Agent LLM Systems with Optimized Prompt Attacks
18:35 - 19:00 (UTC) Discussion