Hamed RAHIMI's Projects
Aligned Neural Topic Model (ANTM) for Exploring Evolving Topics: a dynamic neural topic model that uses document embeddings (data2vec) to compute clusters of semantically similar documents at different periods, and aligns document clusters to represent topic evolution.
INSEAD MBA Course "Building genAI Products and Business", Theodoros Evgeniou
In this project, we aim to use AWS to develop two computing systems within a web application: CloudCal and CloudPic
Citation-informed Neural Topic Models
Contextualised Topic Coherence Metrics: A new way to evaluate neural topic models.
Official Implementation of "DeLLMa: A Framework for Decision Making Under Uncertainty with Large Language Models"
RAG for Discovery
In this Project, we have designed a recommender system that provides geo position of transport facilities such as Train, Bicycle, Bus and Tram in France. Besides, it contains social geo positions such Museums, Hospitals, Universities and Post Offices in France.
Leverage hallucinations from Large Language Models (LLMs) for novelty-driven explorations.
Hybrid Message-Passing Architectures for Node Prediction: uses OGB to explore techniques
🦜🔗 Build context-aware reasoning applications
A multi density clustering algorithm for evolving data stream
Three Agent-Based Simulation for Edge Computing in 5G and Beyond for the recent paper titled "Design and Simulation of a Hybrid Architecture for Edge Computing in 5G and Beyond" by Hamed Rahimi et al.
Implementation of Recent Neural Topic Models (ETM, ATM, Top2Vec, CTM, BERTopic) for 20Newsgroup Corpus.
Optimization using CVXPY python library
LLM Chain for answering questions from documents with citations
SMASH: a Semantic-enabled Multi-agent Approach for Self-adaptation of Human-centered IoT. In this code, we have considered the Smart Home as the case study of smart environments. SMASH agents are provided with a 4-layer architecture based on the BDI agent model that integrates human values with goal-reasoning, planning, and acting. It also takes advantage of a semantic-enabled platform called Home'In to address interoperability issues among non-identical agents and devices with heterogeneous protocols and data formats. This approach is validated by developing a scenario as the proof of concept. The timely responses of SMASH agents show the feasibility of the proposed approach in human-centered environments.
Skeleton-of-Thought: Large Language Models Can Do Parallel Decoding
A collection of topic diversity measures for topic modeling
some old school neural topic models with Pytorch Lightning.
This project is done by Hamed Rahimi for the course web and mobile programming at Mines Saint Etienne
This project is done by Hamed Rahimi for the course web and mobile programming at Mines Saint Etienne