Publications

Activation steering has become a popular way to control Large Language Model (LLM) behavior without fine-tuning. Since the technique is …

Stem retrieval, the task of matching missing stems to a given audio submix, is a key challenge currently limited by models that discard …

Modern deep learning usually treats models as separate artifacts: trained independently, specialized for particular purposes, and …

The Post-AGI Science and Society Workshop at ICLR 2026 asks what comes next if AGI becomes ubiquitous, reliable, and affordable, and …

Reasoning models have demonstrated remarkable capabilities in solving complex problems by leveraging long chains of thought. However, …

The Platonic Representation Hypothesis suggests that independently trained neural networks converge to increasingly similar latent …

Generative audio models, based on diffusion and autoregressive architectures, have advanced rapidly in both quality and expressiveness. …

Foundation models capable of generalizing across species and tasks represent a promising new frontier in bioacoustics, with NatureLM …

Transformer components such as non-linear activations and normalization are inherently non-injective, suggesting that different inputs …

Recurrent-depth transformers scale test-time compute by iterating latent computations before emitting tokens. We study the geometry of …

Task arithmetic has emerged as a simple yet powerful technique for model merging, enabling the combination of multiple finetuned models …

CLIP is a discriminative model trained to align images and text in a shared embedding space. Due to its multimodal structure, it serves …

Model merging allows combining the capabilities of existing models into a new one - post hoc, without additional training. This has …

Foundation models serve as the backbone for numerous specialized models developed through fine-tuning. However, when the underlying …

Recent advances in generative models have made it possible to create high-quality, coherent music, with some systems delivering …

In this work, we introduce an efficient generation procedure to produce synthetic multi-modal datasets of fluid simulations. The …

Model merging has recently emerged as a lightweight alternative to ensembling, combining multiple fine-tuned models into a single set …

Loops–short audio segments designed for seamless repetition–are central to many music genres, particularly those rooted in …

Understanding how large audio models represent music, and using that understanding to steer generation, is both challenging and …

Task Arithmetic has emerged as a simple yet effective method to merge models without additional training. However, by treating entire …

Evolutionary model merging enables the creation of high-performing multi-task models but remains computationally prohibitive for …

Benchmarks are important tools for tracking the rapid advancements in large language model (LLM) capabilities. However, benchmarks are …

Model merging has recently emerged as a cost-efficient paradigm for multi-task learning. Among current approaches, task arithmetic …

In this paper, we present a novel data-free method for merging neural networks in weight space. Differently from most existing works, …

Discover why, when and how distinct learning processes yield similar representations, and the degree to which these can be unified.

Models trained on semantically related datasets and tasks exhibit comparable inter-sample relations within their latent spaces. We …

Datasets used to train deep learning models in industrial settings often exhibit skewed distributions with some samples repeated a …

Getting a good understanding of the user intent is vital for e-commerce applications to surface the right product to a given customer …

Cross-lingual phenomena are quite common in informal contexts like social media, where users are likely to mix their native language …

Few-shot graph classification is a novel yet promising emerging research field that still lacks the soundness of well-established …

Deep learning approaches have recently raised the bar in many fields, from Natural Language Processing to Computer Vision, by …