Poster Sessions: Columbia AI Summit

March 4, 2025

Please note that the floorplans viewed below may be slightly modified due to space restrictions, however the poster number assignment will not change.

Morningside

Location: Low Memorial Library - Faculty Room
Time: 9:00 AM - 10:00 AM

Floorplan:
Morningside Floor Plan
  • P01 - LOW: Brand Prediction through Image-Centric Modeling
  • P02 - LOW: LLM Based Core Value Dossier Creation Agent
  • P03 - LOW: Real-Time Non-Contact Estimation of Running Metrics on Treadmills using Smartphones
  • P04 - LOW: EmbodiedRDA: Connecting Foundation Models with the Physical World using Reconfigurable Drone Agents
  • P05 - LOW: Improving On-Device LLMs' Sensory Understanding with Embedding Interpolations
  • P06 - LOW: SocialMind: LLM-based Proactive AR Social Assistive System with Human-like Perception for In-situ Live Interactions
  • P07 - LOW: SPECTRA: A Drone-based Multispectral Sensing Platform for Complex Environment Perception
  • P08 - LOW: From Conflict to Consensus: Using AI for participatory design and community engagement
  • P09 - LOW: Assessing Graduate Student Attitudes toward ChatGPT and its Effectiveness as a Teaching Tool for Real Estate Finance
  • P10 - LOW: The Impact of AI on the Future of Cities and Urban Living
  • P11 - LOW: Human-AI Synergy in Protein Structure Prediction: Addressing AlphaFold’s Limitations
  • P12 - LOW: Improving Indian Summer Monsoon Rainfall Forecasts at the Subseasonal-to-Seasonal Scale with Deep Learning
  • P13 - LOW: Assessing Functional Relationships for Model Parameters in Ice Sheet Modeling using Deep Learning
  • P14 - LOW: AI-Enhanced Resolution Framework for T2-Weighted MRI with Improved Anatomical Detail
  • P15 - LOW: Diffusion models for determining local atomic structure from X-ray absorption spectra
  • P16 - LOW: From Rewards to Videos: Self-Supervised Temporal Pretraining for Inverse Reinforcement Learning & Video Generative Models
  • P17 - LOW: Particle-Grid Neural Dynamics for Learning Deformable Object Models from Depth Images
  • P18 - LOW: LLM-Powered Compliance Agents:Affordable and Scalable Compliance Solutions for Small- and Medium-sized Businesses (SMBs)
  • P19 - LOW: MallowsPO: Fine-Tune Your LLM with Preference Dispersions
  • P20 - LOW: Machine learning potential for serpentines
  • P21 - LOW: From Experiments to In-Service: Encoding Structural Damage via Speaker Recognition
  • P22 - LOW: Gaze-Informed Vision Transformers: Predicting Driving Decisions Under Uncertainty
  • P23 - LOW: Development of a Generalizable Gradient Boosting Regression Based Framework for Predicting the Properties of Single and Binary IL Systems
  • P24 - LOW: Lessons from the Nuclear Weapons Age for Managing AI Development
  • P25 - LOW: Design- and Run-Time Reconfigurable System-on-chip Architectures for Language Model Inference on the Edge
  • P26 - LOW: Using Expert Gaze for Self-Supervised and Supervised Contrastive Learning of Glaucoma from OCT Data
  • P27 - LOW: AdaptLingo: An Adaptive Speech-to-Speech Conversational Agent Tailored to Users' English Proficiency Levels
  • P28 - LOW: Fast, Accurate Manifold Denoising by Tunneling Riemannian Optimization
  • P29 - LOW: A Hybrid Self-Supervised and Supervised Deep Learning Framework for Accurate Volumetric Breast Density Segmentation From 3D Reconstructed DBT Images
  • P30 - LOW: Commitment to Sparse Strategies in Two-Player Games
  • P31 - LOW: GATO: GPU-Accelerated and Batched Trajectory Optimization for Real-Time Nonlinear Model Predictive Control
  • P32 - LOW: Conformal Uncertainty Quantification of Electricity Price Predictions for Risk-Averse Storage Arbitrage
  • P33 - LOW: ElyonDB: Intelligent Data Discovery and Integration for Large-Scale Analytical Tasks / DAPLab: The Data, Agents, and Processes Lab
  • P34 - LOW: SUSIE: Pharmaceutical ontology-based LLM for accelerated drug development
  • P35 - LOW: Uncertainty Quantification for LLM-Based Survey Simulations
  • P36 - LOW: Linguistic instructions drive flexible behavior in human single neurons
  • P37 - LOW: Mean-Variance Portfolio Selection by Continuous-Time Reinforcement Learning: Algorithms, Regret Analysis, and Empirical Study
  • P38 - LOW: Real-Time Structural Damage Detection Using Kolmogorov-Arnold Neural Networks and Cepstral Features
  • P39 - LOW: In Their Own Words: A Tailored Approach to Measuring Politicized Group Attachments
  • P40 - LOW: From Silicon to Grid: Navigating AI’s Gold Rush in Energy, Dismantling Guardrails and Complexities in Global Regulatory Terrains
  • P41 - LOW: AI That Can Enhance, Not Replace, Student Thinking
  • P42 - LOW: AI, Cultural Art, and Conflict Resolution: Exploring the Role of Technology, Art, Faith, and Adaptive Mediation in Advancing Sustainable Peace
  • P43 - LOW: Bridging AI and Pedagogy: Developing a Customized GenAI TA to Support Metacognitive Engagement in Computing Education
  • P44 - LOW: Exploring the Black Box Hypothesis with AI Assistance
  • P45 - LOW: Public Trust in AI and Institutional Trust in the Era of Digital Transformation
  • P46 - LOW: “It's Not About Laziness, It's About Efficiency”: Youth Perspectives on Generative AI in Higher Education Through the Lens of TikTok
  • P47 - LOW: Make-in-Progress: Multimodal AI for Constructionist Making
  • P48 - LOW: Teaching AI Literacy across Curriculum: Pedagogical Framework for AI Literacy
  • P49 - LOW: Enhancing Education with AI: A RAG-Based Teaching Assistant and Real-Time Conversational Video for Simulation-Based Learning
  • P50 - LOW: Upskilling faculty work with AI

Manhattanville

Location: Lee C. Bollinger Forum - West Atrium
Time: 12:00 PM - 1:00 PM

Floorplan:
Manhattanville Floor Plan
  • P01 - FORUM: Revolutionizing Data-Driven Decisions with xnode: An AI-Powered Business Intelligence Framework Transforming Predictive Analytics and Automation for Enterprise Growth
  • P02 - FORUM: Inductive Bias in AI Models for Physics
  • P03 - FORUM: Large Language Model-Driven Full Lifecycle Architectural Design: A BIM-Based Workflow and Prototype
  • P04 - FORUM: Leveraging T-SNE Visualization of Professional Networks for Predictive AI in Transdisciplinary Team Collaboration
  • P05 - FORUM: From Tiny World to Large World 
  • P06 - FORUM: Imagining the Future: Sociotechnical Imaginaries in Artificial General Intelligence Development at OpenAI
  • P07 - FORUM: Cloudlands: Past, Present, and Futures of Data Center Development
  • P08 - FORUM: What Makes for A Good Thumbnail? Video Content Summarization Into A Single Image 
  • P09 - FORUM: Modeling Dynamic Consumer Preferences with Distribution Shift
  • P10 - FORUM: Consumer Search and Learning Dynamics in E-commerce: An Approximate Dual Dynamic Programming Approach with a Dual Neural Network
  • P11 - FORUM: ENSO's impact on Antarctic sea ice predictability by AI models at subseasonal timescale
  • P12 - FORUM: Leveraging Satellite Remote Sensing Measurements to Build Machine Learning Models to Estimate High-Resolution Gridded PM2.5 in Ghana
  • P13 - FORUM: Mapping visual categorical selectivity across the whole brain using transformer-based encoders and large-scale generative models
  • P14 - FORUM: A Computational Framework for Optimized Firebreak Generation Using Probabilistic Fire Spread Modeling
  • P15 - FORUM: SparseQJL: Efficient KV Cache Compression For Accelerating Large Language Models
  • P16 - FORUM: How does complexity in an infant’s environment shape the way the infant brain responds to complexity in a visual scene?
  • P17 - FORUM: Reimagining Language Learning through the Integration of AI in Spanish Beginner II
  • P18 - FORUM: Leveraging AI for Insights into Corporate Emission Reduction and Climate Action
  • P19 - FORUM: An Adversarial Approach for Humor Generation and Evaluation with LLMs
  • P20 - FORUM: AI-accelerated Personalized Predictive Modeling with Uncertainty for Neonates with Borderline Left Ventricles
  • P21 - FORUM: SLAMuZero: Plan and Learn to Map for Joint SLAM and Navigation
  • P22 - FORUM: Not Just Novelty: A Longitudinal Study on Utility and Customization of an AI Workflow
  • P23 - FORUM: 6G as a Sensor: Repurposing Wireless Signals for AI-Driven Sensing
  • P24 - FORUM: Control of Nonstationary Stochastic Systems with QPLEX
  • P25 - FORUM: NYC Congestion Pricing: A Vision-Based Analysis
  • P26 - FORUM: Probing Adaptive Decision-Making Under Uncertainty Using Extended Hidden Markov Models
  • P27 - FORUM: Towards Neural No-Resource Language Translation: A Comparative Evaluation of Approaches
  • P28 - FORUM: Powering the Next Gold Rush: AI's Energy Supply
  • P29 - FORUM: Aligning AI, Climate, and Energy Leadership: How must diplomacy evolve for humanity to maximize its chances of survival?
  • P30 - FORUM: Understanding Figurative Meaning through Explainable Visual Entailment
  • P31 - FORUM: Causal discovery and inference through next-token prediction
  • P32 - FORUM: Active Exploration via Autoregressive Generation of Missing Data
  • P33 - FORUM: Building Real-World Language Readiness with AI: Leveraging ChatGPT for Job Interview
  • P34 - FORUM: Latent Specificity: Photographic Archives, AI, and LoRAs
  • P35 - FORUM: Re-imagining oral history archives with artificial intelligence
  • P36 - FORUM: Ensuring Fairness in Algorithmic Decision-Making: A Machine Learning Approach to Bias Detection and Error Estimation in Housing Policy – A New York Case Study
  • P37 - FORUM: Writing “right” with multilingual learners: translanguaging, hegemonic writing instruction and the possibility of not 
  • P38 - FORUM: Exploring Neurodivergent Social Media Schema: strategies for teaching, learning, and community building in the post-pandemic era
  • P39 - FORUM: Virtual Reality Exposure Therapy with D-Cycloserine on Anger and 9/11-Related PTSD Symptom Reduction: A Methods Poster
  • P40 - FORUM: AI Awareness Survey of Educators
  • P41 - FORUM: Collective inference of the truth of propositions from crowd probability judgments
  • P42 - FORUM: Categorical abstraction in continuous representations
  • P43 - FORUM: Q-MRS: A Deep Learning Framework for Quantitative Magnetic Resonance Spectra Analysis
  • P44 - FORUM: DriveGenVLM: Real-world Video Generation for Vision Language Model based Autonomous Driving

CUIMC

Location: Vagelos Education Center - Lobby
Time: 2:00 PM - 3:00 PM

Floorplan:
CUIMC Floor Plan
  • P01 - VEC: Long-term Visual Field Appearance Prediction Using Artificial Intelligence for Ophthalmic Education and Patient Follow-up
  • P02 - VEC: Joint Modeling for Learning Decision-Making Dynamics in Behavioral Experiments
  • P03 - VEC: Artificial Intelligence to Detect Spinal Fluid and Spinal Cord on Ultrasound
  • P04 - VEC: Decentralized Personalization for Federated Medical Image Segmentation via Gossip Contrastive Mutual Learning
  • P05 - VEC: Medical-Expert Eye Movement Augmented Vision Transformers for Glaucoma Diagnosis and Ophthalmology Resident Training
  • P06 - VEC: Streamlining radiology operations with machine-learning based musculoskeletal MRI and CT examination protocol selection
  • P07 - VEC: Artificial Intelligence Curriculum for Ophthalmology Residents 
  • P08 - VEC: From computational models to brain networks: Investigating the circuit dynamics underlying working memory
  • P09 - VEC: Navigating Expertise, Culture and Power: Making Impactful Ethics Work in AI Development and Research
  • P10 - VEC: Neural Fine-Gray: Monotonic neural networks for competing risks
  • P11 - VEC: SeqDance: A Protein Language Model for Representing Protein Dynamic Properties
  • P12 - VEC: Mitigating Spurious Invariance in Contrastive Learning: A Probabilistic Self-Supervised Learning Framework for Medical Image Analysis
  • P13 - VEC: Applications of Large Language Models in Healthcare: A Scoping Review
  • P14 - VEC: AI’s Blind Spot: The Unseen Problem of Unreliable AI Provider Suggestions
  • P15 - VEC: Capturing Complexity in Self-Tracked Menstruation: A Foundation Model for Temporal User Data
  • P16 - VEC: OptiRes-DDPM for Portable OCT Image Quality Enhancement Toward Downstream AI-Based Ophthalmic Disease Diagnosis
  • P17 - VEC: Learning from Expert Clinician Gaze: How our Fixation-Order-Informed Vision Transformer's (FOI ViT) Success Exposes Multimodal Medical AI’s Adaptation Gaps
  • P18 - VEC: Automatic DCIS Segmentation in First Post-Contrast DCE-MRI Image using MA-SAM architecture
  • P19 - VEC: Multi-View Transformers for Airway-to-Lung Ratio Estimation from Cardiac CT: The C4R Study
  • P20 - VEC: The Impact of State Space and Dimensionality on Pancreatic Ultrasound Tumor Segmentation
  • P21 - VEC: bneR: A collaborative workflow for air pollution exposure modeling and uncertainty characterization using the Bayesian Nonparametric Ensemble
  • P22 - VEC: Leveraging Generative AI to Identify Social Determinants of Health and Predict Nursing Interventions
  • P23 - VEC: Beyond EHR data: Leveraging Natural Language Processing and ML to Uncover Cognitive Insights from Patient-Nurse Verbal Communications
  • P24 - VEC: Presenting Predictive ML Models to Patients: Lessons from the HeartLogic Pilot Study​
  • P25 - VEC: Eye-Track-ML: A Machine Learning Pipeline for Automated Frame-by-Frame Coding of Eye-Tracking Videos
  • P26 - VEC: Comparing classical and transformer-based NLP for detecting sentiment and first-person language in adolescent social media messages
  • P27 - VEC: Multisite Pragmatic Cluster-Randomized Controlled Trial of the CONCERN Early Warning System 
  • P28 - VEC: Targeting the tumor microenvironment by duvelisib in peripheral T-cell lymphomas
  • P29 - VEC: Enhancing Brain Age Estimation with a Multimodal 3D CNN Approach Combining Structural MRI and AI-Synthesized Cerebral Blood Volume Data