Bruno Bianchi

Bruno Bianchi, Diego Tiscornia, Matias Travizano et al.

Since the advent of Large Language Models (LLMs), a significant area of research has focused on their intrinsic biases, particularly in political discourse. This study investigates a different but related concept, "political plasticity", which is defined as the capacity of models to adapt their responses based on the user supplied context. To analyze this, a testing framework was developed using an expanded corpus of 200 politically-oriented questions across economic and personal freedom axes, based on a prior framework by Lester (1996). The study explored several methods to induce political bias, including simplified and topic-based system prompts, as well as user prompts with few-shot examples. The results show that while system prompts were largely ineffective, user prompts successfully elicited significant ideological shifts, particularly along the Economic Freedom axis in larger and newer models. Through a validation experiment, we examined whether models answer questionnaires by recognizing the underlying question format. Inverting the sense of the questions revealed unexpected, counter-intuitive shifts in most models, suggesting potential data leakage. Finally, we also analyzed how model plasticity varies when the experiment is conducted in different languages. The results reveal subtle yet notable shifts across each of the analyzed languages. Overall, our results indicate that small and older LLMs exhibit limited or unstable political plasticity, whereas newer frontier models display reliable, expected adaptability.

Bruno Bianchi, Alfredo Umfurer, Juan Esteban Kamienkowski

The advancement of the Natural Language Processing field has enabled the development of language models with a great capacity for generating text. In recent years, Neuroscience has been using these models to better understand cognitive processes. In previous studies, we found that models like Ngrams and LSTM networks can partially model Predictability when used as a co-variable to explain readers' eye movements. In the present work, we further this line of research by using GPT-2 based models. The results show that this architecture achieves better outcomes than its predecessors.

Bruno Bianchi, Aakash Agrawal, Stanislas Dehaene et al.

Human readers can accurately count how many letters are in a word (e.g., 7 in ``buffalo''), remove a letter from a given position (e.g., ``bufflo'') or add a new one. The human brain of readers must have therefore learned to disentangle information related to the position of a letter and its identity. Such disentanglement is necessary for the compositional, unbounded, ability of humans to create and parse new strings, with any combination of letters appearing in any positions. Do modern deep neural models also possess this crucial compositional ability? Here, we tested whether neural models that achieve state-of-the-art on disentanglement of features in visual input can also disentangle letter position and letter identity when trained on images of written words. Specifically, we trained beta variational autoencoder ($β$-VAE) to reconstruct images of letter strings and evaluated their disentanglement performance using CompOrth - a new benchmark that we created for studying compositional learning and zero-shot generalization in visual models for orthography. The benchmark suggests a set of tests, of increasing complexity, to evaluate the degree of disentanglement between orthographic features of written words in deep neural models. Using CompOrth, we conducted a set of experiments to analyze the generalization ability of these models, in particular, to unseen word length and to unseen combinations of letter identities and letter positions. We found that while models effectively disentangle surface features, such as horizontal and vertical `retinal' locations of words within an image, they dramatically fail to disentangle letter position and letter identity and lack any notion of word length. Together, this study demonstrates the shortcomings of state-of-the-art $β$-VAE models compared to humans and proposes a new challenge and a corresponding benchmark to evaluate neural models.