Related papers: Hidden Breakthroughs in Language Model Training

Understanding when and how linguistic knowledge emerges during language model training remains a central challenge for interpretability. Most existing tools are post hoc, rely on scalar metrics, or require nontrivial integration effort,…

Research has shown that neural models implicitly encode linguistic features, but there has been no research showing \emph{how} these encodings arise as the models are trained. We present the first study on the learning dynamics of neural…

In-context learning (ICL) enables large language models (LLMs) to acquire new behaviors from the input sequence alone without any parameter updates. Recent studies have shown that ICL can surpass the original meaning learned in pretraining…

Can we teach natural language understanding models to track their beliefs through intermediate points in text? We propose a representation learning framework called breakpoint modeling that allows for learning of this type. Given any text…

When large language models (LLMs) use in-context learning (ICL) to solve a new task, they must infer latent concepts from demonstration examples. This raises the question of whether and how transformers represent latent structures as part…

Data unlearning aims to remove the influence of specific training samples from a trained model without requiring full retraining. Unlike concept unlearning, data unlearning in diffusion models remains underexplored and often suffers from…

The organization of latent knowledge within large-scale models poses unique challenges when addressing overlapping representations and optimizing contextual accuracy. Conceptual redundancies embedded across layers often result in…

With the successful application of deep learning in communications systems, deep neural networks are becoming the preferred method for signal classification. Although these models yield impressive results, they often come with high…

Phase transitions mark qualitative reorganizations of collective behavior, yet identifying their boundaries remains challenging whenever analytic solutions are absent and conventional simulations fail. Here we introduce learnability as a…

The substantial growth of online learning, in particular, Massively Open Online Courses (MOOCs), supports research into the development of better models for effective learning. Learner 'confusion' is among one of the identified aspects…

Decentralized machine learning is a promising emerging paradigm in view of global challenges of data ownership and privacy. We consider learning of linear classification and regression models, in the setting where the training data is…

Unlearning in large language models (LLMs) aims to remove specified data, but its efficacy is typically assessed with task-level metrics like accuracy and perplexity. We show that these metrics can be misleading, as models can appear to…

Understanding the sequence of cognitive operations that underlie decision-making is a fundamental challenge in cognitive neuroscience. Traditional approaches often rely on group-level statistics, which obscure trial-by-trial variations in…

Continual learning studies how models can adapt to new tasks while retaining previously acquired knowledge. Although a broad spectrum of methods has been proposed to mitigate catastrophic forgetting, the field remains predominantly…

Our goal is to understand how post-training methods, such as fine-tuning, alignment, and unlearning, modify language model behavior and representations. We are particularly interested in the brittle nature of these modifications that makes…

Recent advances in theoretical Deep Learning have introduced geometric properties that occur during training, past the Interpolation Threshold -- where the training error reaches zero. We inquire into the phenomena coined Neural Collapse in…

Scaling laws predict loss from compute but not how capabilities interact. We measure the coupling between reasoning and truthfulness across 63 base models from 16 families and find a regime change invisible to loss curves: below a…

This paper presents a method for detecting mispronunciations with the aim of improving Computer Assisted Language Learning (CALL) tools used by foreign language learners. The algorithm is based on Principle Component Analysis (PCA). It is…

This paper presents a novel theoretical framework for understanding how diffusion models can learn disentangled representations. Within this framework, we establish identifiability conditions for general disentangled latent variable models,…

Learners sharing similar implicit cognitive states often display comparable observable problem-solving performances. Leveraging collaborative connections among such similar learners proves valuable in comprehending human learning. Motivated…