14 citations · 43 across the 11 of their papers we have counts for
11 papers
A Causal World Model Underlying Next Token Prediction: Exploring GPT in a Controlled Environment
Raanan Y. Rohekar, Yaniv Gurwicz, Sungduk Yu +2
Are generative pre-trained transformer (GPT) models, trained only to predict the next token, implicitly learning a world model from which sequences are generated one token at a tim…
LVLM-Interpret: An Interpretability Tool for Large Vision-Language Models
Gabriela Ben Melech Stan, Estelle Aflalo, Raanan Yehezkel Rohekar +7
In the rapidly evolving landscape of artificial intelligence, multi-modal large language models are emerging as a significant area of interest. These models, which combine various…
Causal Interpretation of Self-Attention in Pre-Trained Transformers
Raanan Y. Rohekar, Yaniv Gurwicz, Shami Nisimov
We propose a causal interpretation of self-attention in the Transformer neural network architecture. We interpret self-attention as a mechanism that estimates a structural equation…
From Temporal to Contemporaneous Iterative Causal Discovery in the Presence of Latent Confounders
Raanan Y. Rohekar, Shami Nisimov, Yaniv Gurwicz +1
We present a constraint-based algorithm for learning causal structures from observational time-series data, in the presence of latent confounders. We assume a discrete-time, statio…
CLEAR: Causal Explanations from Attention in Neural Recommenders
Shami Nisimov, Raanan Y. Rohekar, Yaniv Gurwicz +2
We present CLEAR, a method for learning session-specific causal graphs, in the possible presence of latent confounders, from attention in pre-trained attention-based recommenders.…
Iterative Causal Discovery in the Possible Presence of Latent Confounders and Selection Bias
Raanan Y. Rohekar, Shami Nisimov, Yaniv Gurwicz +1
We present a sound and complete algorithm, called iterative causal discovery (ICD), for recovering causal graphs in the presence of latent confounders and selection bias. ICD relie…
Improving Efficiency and Accuracy of Causal Discovery Using a Hierarchical Wrapper
Shami Nisimov, Yaniv Gurwicz, Raanan Y. Rohekar +1
Causal discovery from observational data is an important tool in many branches of science. Under certain assumptions it allows scientists to explain phenomena, predict, and make de…
A Single Iterative Step for Anytime Causal Discovery
Raanan Y. Rohekar, Yaniv Gurwicz, Shami Nisimov +1
We present a sound and complete algorithm for recovering causal graphs from observed, non-interventional data, in the possible presence of latent confounders and selection bias. We…
Modeling Uncertainty by Learning a Hierarchy of Deep Neural Connections
Raanan Y. Rohekar, Yaniv Gurwicz, Shami Nisimov +1
Modeling uncertainty in deep neural networks, despite recent important advances, is still an open problem. Bayesian neural networks are a powerful solution, where the prior over ne…
Bayesian Structure Learning by Recursive Bootstrap
Raanan Y. Rohekar, Yaniv Gurwicz, Shami Nisimov +2
We address the problem of Bayesian structure learning for domains with hundreds of variables by employing non-parametric bootstrap, recursively. We propose a method that covers bot…
Constructing Deep Neural Networks by Bayesian Network Structure Learning
Raanan Y. Rohekar, Shami Nisimov, Yaniv Gurwicz +2
We introduce a principled approach for unsupervised structure learning of deep neural networks. We propose a new interpretation for depth and inter-layer connectivity where conditi…