Impact of AML Scoring Granularity on Elliptic++ Graph Analysis
Explore how transaction-level vs actor-level AML scoring affects investigation outcomes using the Elliptic++ Bitcoin dataset.
Agentic Adversarial Attacks Reveal NLP Pipeline Weaknesses
Discover how agentic adversarial rewriting exposes key vulnerabilities in black-box NLP pipelines, highlighting risks and defense strategies.
Escher-Loop: Adaptive Evolution for Autonomous Agents
Discover Escher-Loop, a closed-loop framework enabling mutual evolution of task and optimizer agents for smarter autonomous AI systems.
Detecting Misaligned Reasoning in Continuous Thought AI Models
Explore methods to detect misaligned reasoning in continuous thought AI models using the MoralChain benchmark for safer, ethical AI development.
ArguAgent: AI-Driven Real-Time Grouping for STEM Debate
Discover how ArguAgent uses AI to create real-time student groups, boosting productive argumentation and engagement in STEM classrooms.