ANAIS 2026 Program Schedule to be live soon.
Foundational concepts in ML
Geometric Deep Learning (GDL)
Neuro-symbolic AI
World Models
Generative Modeling
Applications of AI in healthcare, finance, agriculture, and more)
The 11-day school will start on 28th December, 2026, with a day of hiking on 1st January 2027.
Daily Schedule: 08:30 to 18:00
Lectures (~ 45 hrs)
Hands-on Practical Labs (~ 12 hrs)
Poster Presentation
Project Sessions
Lunch Break: 1 hour 30 minutes
Tea Breaks: Two 30-minute breaks
Total Instructional Time: Approximately 7 hours per day
Last Year's Program Schedule
The program will also feature a dedicated Poster Presentation Session, providing participants with an opportunity to showcase their ongoing research and scholarly work. The session is designed to promote knowledge exchange, encourage idea sharing, and provide constructive feedback from mentors, peers, and fellow researchers.
Selected participants will be invited to present their work through posters, followed by interactive discussions and networking opportunities.
The Poster Presentation Session is open only to registered ANAIS participants. Details regarding poster submission, selection, and presentation guidelines will be provided during the registration process.
The following 17 posters were presented during the Poster Presentation Session.
Adiabatic Reinforcement Learning
AI-driven System For Post-disaster Agricultural Damage Assessment
An AI-Enabled IoT Framework for Forest Fire Detection and Alerting with Hybrid Connectivity
Beyond SMOTE: Improving Credit Card Fraud Detection on Imbalanced Data with Conditional GANs and XGBoost
Deep Learning–Based Disease Detection in Cauliflower Using YOLOv8
Evaluating In-domain Transfer Learning for Hippocampus Segmentation in T1-weighted MR Images
FLDWM: A world model approach for de novo molecule generation
Gradient Optimisation Algorithms for (L0-L1)-smooth Functions => Best Theoretical Poster
Learning Forecast Bias: Climate Offset Correction Using NASA Reanalysis and Sparse Station Data in Nepal
MathPrereq: Dynamic Prerequisite Knowledge Identification using Large Language Models and Knowledge Graphs in Foundational Mathematics
Multi-stream Physics Hybrid Networks for solving Navier-Stokes equations
NepScript Genesis: Neural Architecture Search for Devanagari Handwritten Digit Synthesis
SiDGen: Structure-informed Diffusion for Generative Modelling of Ligands for Proteins
SigTem: A Non-Invertible Technique for Online Signature Template Protection => Best Application Poster
Stage-aware Crop Advisory System With Adaptive Cropping Calendar Adjustment
Swarlekha: Auto-regressive Zero-shot Bilingual Voice Cloning and Text-to-speech model
Towards Nepali-language LLMs: Efficient GPT training with a Nepali BPE tokeniser
The Project Session encourages networking, collaborative thinking, innovation, and the practical application of AI to real-world challenges. Participants will work in groups to discuss, develop, and propose AI-based project ideas.
Projects may focus on AI research or practical applications, with emphasis on problem relevance, innovation, feasibility, and potential social or industrial impact. Participants are encouraged to collaborate actively with their teammates and mentors and make the most of this opportunity for hands-on learning, idea development, and networking.
More details about the format and activities of the Project Session will be provided later.
Hiking Details
DESTINATION: TBA
DATE: 1st Jan, 2027
DAY: Thursday
MEET-UP TIME: 8 am
MEETING LOCATION: TBA
The program from the 2018 ANAIS Winter Edition is available here.
The program from the 2019 ANAIS Winter Edition is available here.
The program from the 2021 ANAIS Winter Edition is available here.
The program from the 2023 ANAIS Spring Edition is available here.
The program from the 2024 ANAIS Winter Edition is available here.
The program from the 2025 ANAIS Winter Edition is available here.
The video recordings from the 2021 ANAIS Winter Edition are available here.
The video recordings from the 2018 ANAIS Winter Edition are available here.