Soran Ghaderi

Master's Student in AI at the University of Essex

About Me

About

I am a Master's student pursuing a specialization in Artificial Intelligence at the University of Essex, Under the supervision of Professor Luca Citi. Previously, I obtained my bachelor's degree from the University of Kurdistan. I have experience in developing deep learning pipelines through contributions to the open-source community and research collaborations.


Research Interests

My ultimate objective is to study the cognitive mechanisms underlying intelligence and develop agents able to reason and interact with the real world.

A major challenge in current AI models is their struggle with consistent logical reasoning and decision-making, especially in scenarios that require multi-step inference or handling of previously unseen situations. Current LLMs can generate highly semantically coherent text, yet, often fail in robust reasoning abilities that are necessary for tasks such as complex problem-solving, planning, or adapting to novel environments.

To this end, my research explores fully differentiable approaches for implementing multi-step reasoning in LLMs, decision-making, zero-shot learning within uncertain environments, and embodied agents. This involves developing new architectures that can maintain coherent thought processes across multiple inference steps.

In a wider context, I am fascinated by the idea of developing a network of specialized networks such as memory and rapid learning, goal-driven planning, spatial reasoning, and error detection and conflict monitoring that operate collaboratively to consistently make good decisions within a complex environment. I am interested in transformers and attention mechanisms with a bias toward efficient computation in these settings: generative models, multimodal learning, and self-supervised learning.



Awards

  • Full funding award for postgraduate studies in Artificial Intelligence, University of Essex

Projects

TransformerX

2021-2022
A Python library for building transformer-based models.

Keywords -- Deep-learning, Machine-learning, Transformers, Attention mechanisms, Algorithms, python library
PyPI - Python Version PyPI - Implementation GitHub last commit PyPI - Maintenance PyPI - License PyPI - Format

Emgraph

2021-2022
A Python toolkit for embedding knowledge graph, developing and evaluating models

Keywords -- Deep-learning, Machine-learning, Graph, Knowledge-graph-embedding, Algorithms
GitHub Repo stars PyPI - Downloads latest release Tensorflow 2

Bigraph

2019-2022
A python library for link prediction in bipartite-networks

Keywords -- link-prediction, Machine-learning, Graph, Graph-analysis
GitHub Repo stars PyPI - Downloads latest release

TASE (Telegram Audio Search Engine)

2020-2022
A lightening-fast audio search engine python library on top of the Telegram messenger platform

Keywords -- Audio-search-engine, Audio-indexing, Knowledge-graph
GitHub Repo stars PyPI - Downloads