Publications

Tracing Stereotypes in Pre-trained Transformers: From Biased Neurons to Fairer Models
Gianmario Voria
Moses Openja
Foutse Khomh
Gemma Catolino
Fabio Palomba
The advent of transformer-based language models has reshaped how AI systems process and generate text. In software engineering (SE), these m… (see more)odels now support diverse activities, accelerating automation and decision-making. Yet, evidence shows that these models can reproduce or amplify social biases, raising fairness concerns. Recent work on neuron editing has shown that internal activations in pre-trained transformers can be traced and modified to alter model behavior. Building on the concept of knowledge neurons, neurons that encode factual information, we hypothesize the existence of biased neurons that capture stereotypical associations within pre-trained transformers. To test this hypothesis, we build a dataset of biased relations, i.e., triplets encoding stereotypes across nine bias types, and adapt neuron attribution strategies to trace and suppress biased neurons in BERT models. We then assess the impact of suppression on SE tasks. Our findings show that biased knowledge is localized within small neuron subsets, and suppressing them substantially reduces bias with minimal performance loss. This demonstrates that bias in transformers can be traced and mitigated at the neuron level, offering an interpretable approach to fairness in SE.
2026-01-08
ArXiv (preprint)
doi.org
arxiv.org
Large language models can effectively convince people to believe conspiracies
Thomas H Costello
Kellin Pelrine
Matthew Kowal
Antonio A. Arechar
Jean-François Godbout
Adam Gleave
David G. Rand
Gordon Pennycook
Large language models (LLMs) have been shown to be persuasive across a variety of contexts. But it remains unclear whether this persuasive p… (see more)ower advantages truth over falsehood, or if LLMs can promote misbeliefs just as easily as refuting them. Here, we investigate this question across three pre-registered experiments in which participants (N = 2,724 Americans) discussed a conspiracy theory they were uncertain about with GPT-4o, and the model was instructed to either argue against ("debunking") or for ("bunking") that conspiracy. When using a"jailbroken"GPT-4o variant with guardrails removed, the AI was as effective at increasing conspiracy belief as decreasing it. Concerningly, the bunking AI was rated more positively, and increased trust in AI, more than the debunking AI. Surprisingly, we found that using standard GPT-4o produced very similar effects, such that the guardrails imposed by OpenAI did little to prevent the LLM from promoting conspiracy beliefs. Encouragingly, however, a corrective conversation reversed these newly induced conspiracy beliefs, and simply prompting GPT-4o to only use accurate information dramatically reduced its ability to increase conspiracy beliefs. Our findings demonstrate that LLMs possess potent abilities to promote both truth and falsehood, but that potential solutions may exist to help mitigate this risk.
2026-01-07
ArXiv (preprint)
doi.org
arxiv.org
Learning Latent Action World Models In The Wild
Quentin Garrido
Tushar Nagarajan
Basile Terver
Nicolas Ballas
Yann Lecun
Michael G. Rabbat
Agents capable of reasoning and planning in the real world require the ability of predicting the consequences of their actions. While world … (see more)models possess this capability, they most often require action labels, that can be complex to obtain at scale. This motivates the learning of latent action models, that can learn an action space from videos alone. Our work addresses the problem of learning latent actions world models on in-the-wild videos, expanding the scope of existing works that focus on simple robotics simulations, video games, or manipulation data. While this allows us to capture richer actions, it also introduces challenges stemming from the video diversity, such as environmental noise, or the lack of a common embodiment across videos. To address some of the challenges, we discuss properties that actions should follow as well as relevant architectural choices and evaluations. We find that continuous, but constrained, latent actions are able to capture the complexity of actions from in-the-wild videos, something that the common vector quantization does not. We for example find that changes in the environment coming from agents, such as humans entering the room, can be transferred across videos. This highlights the capability of learning actions that are specific to in-the-wild videos. In the absence of a common embodiment across videos, we are mainly able to learn latent actions that become localized in space, relative to the camera. Nonetheless, we are able to train a controller that maps known actions to latent ones, allowing us to use latent actions as a universal interface and solve planning tasks with our world model with similar performance as action-conditioned baselines. Our analyses and experiments provide a step towards scaling latent action models to the real world.
2026-01-07
ArXiv (preprint)
doi.org
arxiv.org
Path Following Guidance Strategy for Autonomous Dynamic Soaring
Zihao Zhuo
Meyer Nahon
Inna Sharf
Dynamic soaring is a flight mode that harvests energy from the vertical gradient of horizontal wind and can be used to increase the enduranc… (see more)e and range of unmanned aerial vehicles. Previous studies have mainly focused on trajectory optimization for dynamic soaring, while the problem of following these optimal paths with an autonomous glider has received limited attention. This study proposes a novel guidance strategy that enables precise tracking of an optimal dynamic soaring path on an autonomous glider vehicle. The proposed guidance strategy combines a geometric path-following guidance law with a command projection module specifically designed to address the unique challenges of dynamic soaring, such as the presence of crosswind components and the underactuated nature of glider vehicles. The performance of the proposed guidance strategy is demonstrated through numerical simulations of a 2 m wingspan glider executing dynamic soaring maneuvers in both ridge and surface wind shear layers.
2026-01-07
AIAA Science and Technology Forum and Exposition (published)
doi.org
Evolving Programmatic Skill Networks
Haochen Shi
Xingdi Yuan
Bang Liu
We study continual skill acquisition in open-ended embodied environments where an agent must construct, refine, and reuse an expanding libra… (see more)ry of executable skills. We introduce the Programmatic Skill Network (PSN), a framework in which skills are executable symbolic programs forming a compositional network that evolves through experience. PSN defines three core mechanisms instantiated via large language models: (1)REFLECT for structured fault localization over skill compositions, (2) progressive optimization with maturity-aware update gating that stabilizes reliable skills while maintaining plasticity for uncertain ones, and (3) canonical structural refactoring under rollback validation that maintains network compactness. We further show that PSN's learning dynamics exhibit structural parallels to neural network training. Experiments on MineDojo and Crafter demonstrate robust skill reuse, rapid adaptation, and strong generalization across open-ended task distributions.\footnote{We plan to open-source the code.
2026-01-06
arXiv (preprint)
doi.org
arxiv.org
Genetic and Causal Insights Into White Matter Hyperintensities Across the Brain‐Body Axis
Manpreet Singh
Kimia Shafighi
Flavie E. Detcheverry
Gabrielle Dagasso
Fanta Dabo
Ikrame Housni
Sridar Narayanan
Nils D. Forkert
Sarah A Gagliano Taliun
Danilo Bzdok
AmanPreet Badhwar
White matter hyperintensities (WMHs), visible as bright regions on T2‐weighted FLAIR MRI, are frequent with age and elevated in Alzheimer'… (see more)s disease (AD). Representing axonal damage, demyelination, and edema, WMHs are driven by vascular mechanisms, including endothelial dysfunction and impaired cerebrovascular autoregulation. WMHs also exhibit strong heritability (55–73%), with overlapping genetic pathways shared with AD. Emerging evidence suggests systemic factors across the brain‐body axis influence WMHs, yet these contributions and their genetic overlap with AD remain underexplored. Our study investigated genetic underpinnings specific to WMHs and those shared with AD by assessing partitioned heritability of WMHs and AD across the brain‐body axis with SNP level tissue‐ and cell‐specific annotations; identifying genes associated with WMHs and AD through integration of gene expression data, establishing causal links between SNP‐level findings and imaging‐derived phenotypes (IDPs), particularly structural variations in regional brain volumes. Partitioned heritability was assessed using stratified‐linkage disequilibrium score regression (sLDSC) on GWAS summary statistics ( N  = 3 WMH studies; N  = 6 AD studies) using human A1) tissue level annotations ( N  = 10) and A2) continuous cell‐specific annotations ( N  = 64). MAGMA and FUSION analyses highlighted genes associated with WMH and AD for further bioinformatics analysis (using human protein atlas (HPA) and STRING database). MACAW (Vigneshwaran et al, 2024) modeled causal relationships between WMH‐associated SNPs (from FUMA analysis) and IDPs ( N  = 172), leveraging directed acyclic graphs to evaluate genetic effects while controlling for confounders (Figure 2). Tissue‐specific analysis revealed significant enrichment of WMH‐associated SNPs in the CNS, liver, cardiovascular system, and kidneys, while AD‐associated SNPs were enriched in the CNS, connective bone, liver, and immune tissues. (Figure 1). Cell‐specific analysis identified vascular endothelial cells as enriched across WMH‐enriched tissues. MAGMA analysis, combined with HPA analysis, corroborated sLDSC tissue‐level findings. MAGMA and FUSION analyses highlighted genes associated with WMHs ( N  = 39 and 69) and AD ( N  = 291 and 193). MACAW linked WMH‐associated SNP to 172 IDPs, consistently impacting WM hypointensities and regional brain volumes (e.g., left inferior temporal volume). Our findings highlight systemic multi‐tissue contributions (CNS, liver, cardiovascular system, and kidneys) to WMHs, driven by vascular endothelial dysfunction and shared AD genetics, with SNPs across the body also affecting brain imaging derived phenotypes.
2026-01-06
Alzheimer's & Dementia (published)
doi.org
Preface
Tal Arbel
Ismail Ben Ayed
Marleen de Bruijne
Maxime Descoteaux
Hervé Lombaert
Christopher Pal
2026-01-06
Acta Neerlandica (published)
doi.org
proceedings.mlr.press
Scalable Tree Ensemble Proximities in Python
Adrien Aumon
Guy Wolf
Kevin R. Moon
Jake S. Rhodes
2026-01-05
ArXiv (preprint)
doi.org
arxiv.org
Seeing the forest and the trees: a workflow for automatic acquisition of ultra-high resolution drone photos of tropical forest canopies to support botanical and ecological studies
Étienne Laliberté
Antoine Caron-Guay
Vincent Le Falher
Guillaume Tougas
Helene C. Muller-Landau
Gonzalo Rivas-Torres
Thomas R. Walla
Hugo Baudchon
Melvin Hernández
Adrian Buenaño
Anna Weber
Jeffrey Q. Chambers
Jomber Chota Inuma
Fernando Araúz
Jorge Valdes
Andrés Hernández
David Brassfield
P. Sérgio
Vicente Vasquez
Adriana Simonetti … (see 7 more)
Daniel Magnabosco Marra
Caroline de Moura Vasconcelos
Jarol Fernando Vaca
Geovanny Rivadeneyra
José Illanes
Luis A. Salagaje-Muela
Jefferson Gualinga
Tropical forest canopies contain many tree and liana species, and foliar and reproductive characteristics useful for taxonomic identificatio… (see more)n are often difficult to see from the forest floor. As such, taxonomic identification often becomes a bottleneck in tropical forest inventories. Here we present a drone-based workflow to automatically acquire large volumes of close-up, ultra-high resolution photos of selected tree crowns (or specific locations over the canopy) to support tropical botanical and ecological studies ( https://youtu.be/80goMEifpc4 ). Our workflow is built around the small, easy-to-use DJI Mavic 3 Enterprise (M3E) drone, which is equipped with a wide-angle and a telephoto camera. On day one, the pilot maps a forest area of up to ∼200 ha with the wide-angle camera to generate a high-resolution digital surface model (DSM) and orthomosaic using structure-from-motion (SfM) photogrammetry. On subsequent days, the pilot acquires close-up photos with the telephoto camera from up to 300 selected canopy trees per day. These close-up photos are acquired from 6 m above the canopy and contain a high level of visual detail that allows botanists to reliably identify many tree and liana species. The photos are geolocated with survey-grade accuracy using RTK GNSS, thus facilitating spatial co-registration with other data sources, including the photogrammetry products. The primary operational challenge of our workflow is the need to maintain RTK corrections with the drone to ensure that close-up photos are acquired exactly at the predefined locations. The maximum operational range we achieved was 3 km, which would allow the pilot to reach any tree within a ∼2800 ha area from the take-off point. Although our workflow was developed to support taxonomic identification of tropical trees and lianas, it could be extended to any other forest or vegetation type to support botanical, phenological, and ecological studies. We provide harpia , an open-source Python library to program these automatic close-up photo missions with the M3E drone ( https://github.com/traitlab/harpia ). We provide harpia , an open-source Python library to program these automatic close-up photo missions ( https://github.com/traitlab/harpia ). Drone imagery and labelled close-up photo data are not yet publicly available because they were acquired with the goal of publishing benchmark machine learning datasets and models for tree and liana species classification and prior publication of the data would jeopardize this future publication.
2026-01-05
bioRxiv (preprint)
doi.org
Diffusion tractography outside the brain: the road less travelled
Kurt G. Schilling
Irvin Teh
Julien Cohen-Adad
Richard Dortch
Ibrahim Ibrahim
Nian Wang
Bruce Damon
Rory L. Cochran
Alexander Leemans
Diffusion tractography is a powerful MRI technique for mapping fibrous tissue architecture, traditionally applied to the white matter of the… (see more) brain. This report surveys the growing application of tractography to anatomical structures outside the brain, a domain that presents both unique challenges and unique opportunities. We examine its use in the heart, spinal cord, peripheral nerves, brachial plexus, kidney, skeletal muscle, and prostate. For each region, we detail the necessary methodological adaptations for acquisition, modeling, and processing, and highlight the unique anatomical information that can be derived for research and clinical applications. While significant challenges remain - spanning technical hurdles like physiological motion and susceptibility artifacts, to biological complexities like lower anisotropy and the interpretation of streamline validity - tractography beyond the brain provides invaluable, non-invasive insights into tissue micro-organization, opening a new frontier for biomedical imaging.
2026-01-04
Brain Structure & Function (published)
doi.org
Protein Language Models: Is Scaling Necessary?
Quentin Fournier
Robert M. Vernon
Almer Van Der Sloot
Benjamin Schulz
Sarath Chandar
Christopher James Langmead
Public protein sequence databases contain samples from the fitness landscape explored by nature. Protein language models (pLMs) pre-trained … (see more)on these sequences aim to capture this landscape for tasks like property prediction and protein design. Following the same trend as in natural language processing, pLMs have continuously been scaled up. However, the premise that scale leads to better performance assumes that source databases provide an accurate representation of the underlying fitness landscape, which is likely false. By developing an efficient codebase, designing a modern architecture, and addressing data quality concerns such as sample bias, we introduce AMPLIFY, a best-in-class pLM that is orders of magnitude less expensive to train and deploy than previous models. Furthermore, to support the scientific community and democratize the training of pLMs, we have open-sourced AMPLIFY’s pre-training codebase, data, and model checkpoints.
2026-01-04
BioRxiv (preprint)
doi.org
Shaped by meaning, weighted by reliability: New insights into multisensory integration
Elizaveta Sycheva
Léa St-Gelais
Karim Jerbi CoCo Lab
Shahab Bakhtiari
Franco Lepore
Vanessa Hadid
2026-01-04
bioRxiv (preprint)
doi.org