Are LLM Belief Updates Consistent with Bayes’ Theorem?
This paper evaluates whether large language models (LLMs) update their beliefs over propositions in a Bayesian-consistent manner as they scale in size and capability.
Portfolio
AI Agent for Financial Trend Forecasting with bunq API
Built an AI agent in 24 hours to analyze user banking data, predict financial trends, and create meme-based insights—Top 3 project at the bunq Hackathon.
Symbolic, Subsymbolic, and Hybrid Approaches in AI: A Philosophical Perspective
Exploring the ongoing debate in AI between symbolic, subsymbolic, and hybrid approaches, with a focus on deep learning limitations and the promise of neuro-compositional systems.
Testing Machine Intelligence: Mechanistic Interpretability and Commonsense Reasoning
Exploring commonsense reasoning as a cornerstone of machine intelligence and evaluating mechanistic interpretability as a method for assessing AI capabilities.
ERA: A Goal-Based Agent Integrating LLMs and Ontology for Fake News Detection
Analyzing the ERA agent’s architecture, functionalities, and its applications in determining the veracity of news headlines
(RL) Exploration of Reinforcement Learning Algorithms
Implementing Epsilon-Greedy, Semi-Grad SARSA, and Q-Learning for solving multi-armed bandit and episodic MDP tasks
(NLP) Probing BERT: Understanding Language Model Behavior
Exploring the interpretability of BERT’s outputs through probing techniques for linguistic and semantic analysis
(CI) Predicting Intervention Results & PC Algorithm
Exploring structural causal models and the PC algorithm for causal discovery
(NLP) Exploring LSTM for UPOS Tagging and BERT for Word Sense Disambiguation
Evaluating sequence labeling and contextual embeddings for linguistic tasks
(NLP) Exploring Parsing Algorithms: Viterbi and CKY
Analyzing probabilistic parsing techniques for syntactic structure recognition
(NLP) Analyzing N-Grams and Perplexity for Language Modeling
A study on the effectiveness of n-gram models and perplexity evaluation for language prediction tasks
(XAI) Explaining Readmission of Diabetes Patients
Analyzing factors contributing to diabetes patient readmissions using various explainable AI techniques
(XAI) A Study on Personality Predictions
Exploring various models for job interview invitation predictions using explainability techniques
MEG Data Analysis using Deep Learning
Exploring VisionTransformers and RNNs for brain state classification
(NLP) Opinion Spam Detection using Machine Learning
A study on classifying deceptive reviews using Naive Bayes, Logistic Regression, Decision Trees, and Random Forests
Handwritten Digit Classification using Logistic Regression and SVM
A comparative analysis between logistic regression and SVM models for digit recognition tasks
3D Voxel-based Person Tracking
Offline color modeling and real-time voxel labeling for multi-person tracking
(NLP) Glove
Implementation of small scale variant of the paper GloVe
Numerical and Analytical methods for spatially-extended neurobiological networks
Implementation of tutorials from a neuroscience masterclass for the CCSS at the Utrecht University
Action Recognition in Still Images and Videos
Implementation of a Two-stream network for classifying actions
CNN Architecture Optimization
Implementation and optimization of LeNet variants for FashionMNIST Classification