Ashia Wilson
3 papers · Latest:
Artificial Intelligence
Alignment has a Fantasia Problem
AI often fails when users' goals are unformed; this paper rethinks alignment to help users refine intent, bridging ML, design, and behavioral science.
2604.21827
Human-Computer InteractionCreo: From One-Shot Image Generation to Progressive, Co-Creative Ideation
Creo is a multi-stage text-to-image system that enhances user control and creativity by progressing from sketches to high-res outputs with incremental changes.
2604.13956
Machine LearningOnline Reasoning Calibration: Test-Time Training Enables Generalizable Conformal LLM Reasoning
ORCA uses test-time training and conformal prediction to calibrate LLM reasoning, significantly boosting efficiency and generalization while providing valid confidence.
2604.01170
📬 Weekly AI Paper Digest
Get the top 10 AI/ML arXiv papers from the week — summarized, scored, and delivered to your inbox every Monday.