arXiv subjects

Freya Behrens

Publications and source records attributed to Freya Behrens.

14 recordsLinked to original sources

arXiv2025

Minority Takeover in Majority Dynamics: Searching for Rare Initializations via the History Passing Algorithm

We investigate how much bias in the initial configuration is required to drive global agreement in synchronous, deterministic majority dynamics on large random $d$-regular graphs. Nodes take values $\pm 1$ and update their states at each discrete time step to align with the majority of their neighbors. Using the backtracking dynamical cavity method (BDCM), we estimate the minimal fraction of initial $+1$ nodes required to achieve a $+1$ consensus in $p$ time steps. Our analysis predicts that for $d\geq4$ an initial global minority of $+1$ nodes is sufficient to quickly steer the entire system toward consensus on $+1$. We then investigate whether such initial conditions can be determined explicitly for a given large random regular graph. To this end, we introduce a new algorithm, which we name history-passing reinforcement (HPR), designed to find such initial configurations with a minority of $+1$ nodes. We find, as a main result, that the HPR algorithm finds initial configurations where the minority takes over the majority for $d$-regular random graphs with $d\geq4$. The HPR algorithm outperforms standard simulated annealing-based methods, but does not reach the lowest densities predicted by the BDCM. Rather, the lowest density achievable by the algorithm is near the onset of a dynamical one-step replica symmetry breaking (d1RSB) phase, which we estimate using a one-step replica symmetry breaking (1RSB) formulation of the BDCM. While we focus on the majority dynamics and random $d$-regular graphs, the algorithm can be extended to other dynamical rules and classes of sparse graphs.

cond-mat.dis-nn↗

arXiv2025

Dataset distillation for memorized data: Soft labels can leak held-out teacher knowledge

Dataset distillation aims to compress training data into fewer examples via a teacher, from which a student can learn effectively. While its success is often attributed to structure in the data, modern neural networks also memorize specific facts, but if and how such memorized information is can transferred in distillation settings remains less understood. In this work, we show that students trained on soft labels from teachers can achieve non-trivial accuracy on held-out memorized data they never directly observed. This effect persists on structured data when the teacher has not generalized.To analyze it in isolation, we consider finite random i.i.d. datasets where generalization is a priori impossible and a successful teacher fit implies pure memorization. Still, students can learn non-trivial information about the held-out data, in some cases up to perfect accuracy. In those settings, enough soft labels are available to recover the teacher functionally - the student matches the teacher's predictions on all possible inputs, including the held-out memorized data. We show that these phenomena strongly depend on the temperature with which the logits are smoothed, but persist across varying network capacities, architectures and dataset compositions.

cs.LG↗

arXiv2024

Dynamical Cavity Method for Hypergraphs and its Application to Quenches in the k-XOR-SAT Problem

The dynamical cavity method and its backtracking version provide a powerful approach to studying the properties of dynamical processes on large random graphs. This paper extends these methods to hypergraphs, enabling the analysis of interactions involving more than two variables. We apply them to analyse the $k$-XOR-satisfiability ($k$-XOR-SAT) problem, an important model in theoretical computer science which is closely related to the diluted $p$-spin model from statistical physics. In particular, we examine whether the quench dynamics -- a deterministic, locally greedy process -- can find solutions with only a few violated constraints on $d$-regular $k$-uniform hypergraphs. Our results demonstrate that the methods accurately characterize the attractors of the dynamics. It enables us to compute the energy reached by typical trajectories of the dynamical process in different parameter regimes. We show that these predictions are accurate, including cases where a classical mean-field approach fails.

cond-mat.dis-nn↗

arXiv2024

Could ChatGPT get an Engineering Degree? Evaluating Higher Education Vulnerability to AI Assistants

AI assistants are being increasingly used by students enrolled in higher education institutions. While these tools provide opportunities for improved teaching and education, they also pose significant challenges for assessment and learning outcomes. We conceptualize these challenges through the lens of vulnerability, the potential for university assessments and learning outcomes to be impacted by student use of generative AI. We investigate the potential scale of this vulnerability by measuring the degree to which AI assistants can complete assessment questions in standard university-level STEM courses. Specifically, we compile a novel dataset of textual assessment questions from 50 courses at EPFL and evaluate whether two AI assistants, GPT-3.5 and GPT-4 can adequately answer these questions. We use eight prompting strategies to produce responses and find that GPT-4 answers an average of 65.8% of questions correctly, and can even produce the correct answer across at least one prompting strategy for 85.1% of questions. When grouping courses in our dataset by degree program, these systems already pass non-project assessments of large numbers of core courses in various degree programs, posing risks to higher education accreditation that will be amplified as these models improve. Our results call for revising program-level assessment design in higher education in light of advances in generative AI.

cs.CY↗

arXiv2024

Counting in Small Transformers: The Delicate Interplay between Attention and Feed-Forward Layers

Next to scaling considerations, architectural design choices profoundly shape the solution space of transformers. In this work, we analyze the solutions simple transformer blocks implement when tackling the histogram task: counting items in sequences. Despite its simplicity, this task reveals a complex interplay between predictive performance, vocabulary and embedding sizes, token-mixing mechanisms, and feed-forward layer capacity. We identify two theoretical counting strategies transformers adopt, relation-based and inventory-based counting, each defining distinct learning regimes for the task. These strategies dictate how functionality is distributed between attention and feed-forward layers. We further show that adding softmax and beginning-of-sequence tokens allow for more robustness when embedding dimensions are comparatively small. Empirical introspection of trained models closely confirms both the learning regimes of the various architectures and the formation of these strategies during training. We demonstrate how a basic task that requires only aggregation and selection is significantly impacted by minor design changes.

cs.LG↗

arXiv2024

Counting and Hardness-of-Finding Fixed Points in Cellular Automata on Random Graphs

We study the fixed points of outer-totalistic cellular automata on sparse random regular graphs. These can be seen as constraint satisfaction problems, where each variable must adhere to the same local constraint, which depends solely on its state and the total number of its neighbors in each possible state. Examples of this setting include classical problems such as independent sets or assortative/dissasortative partitions. We analyse the existence and number of fixed points in the large system limit using the cavity method, under both the replica symmetric (RS) and one-step replica symmetry breaking (1RSB) assumption. This method allows us to characterize the structure of the space of solutions, in particular, if the solutions are clustered and whether the clusters contain frozen variables. This last property is conjectured to be linked to the typical algorithmic hardness of the problem. We bring experimental evidence for this claim by studying the performance of the belief-propagation reinforcement algorithm, a message-passing-based solver for these constraint satisfaction problems.

cond-mat.dis-nn↗

arXiv2024

Quenches in the Sherrington-Kirkpatrick model

The Sherrington-Kirkpatrick (SK) model is a prototype of a complex non-convex energy landscape. Dynamical processes evolving on such landscapes and locally aiming to reach minima are generally poorly understood. Here, we study quenches, i.e. dynamics that locally aim to decrease energy. We analyse the energy at convergence for two distinct algorithmic classes, single-spin flip and synchronous dynamics, focusing on greedy and reluctant strategies. We provide precise numerical analysis of the finite size effects and conclude that, perhaps counter-intuitively, the reluctant algorithm is compatible with converging to the ground state energy density, while the greedy strategy is not. Inspired by the single-spin reluctant and greedy algorithms, we investigate two synchronous time algorithms, the sync-greedy and sync-reluctant algorithms. These synchronous processes can be analysed using dynamical mean field theory (DMFT), and a new backtracking version of DMFT. Notably, this is the first time the backtracking DMFT is applied to study dynamical convergence properties in fully connected disordered models. The analysis suggests that the sync-greedy algorithm can also achieve energies compatible with the ground state, and that it undergoes a dynamical phase transition.

cond-mat.dis-nn↗

arXiv2024

A phase transition between positional and semantic learning in a solvable model of dot-product attention

Many empirical studies have provided evidence for the emergence of algorithmic mechanisms (abilities) in the learning of language models, that lead to qualitative improvements of the model capabilities. Yet, a theoretical characterization of how such mechanisms emerge remains elusive. In this paper, we take a step in this direction by providing a tight theoretical analysis of the emergence of semantic attention in a solvable model of dot-product attention. More precisely, we consider a non-linear self-attention layer with trainable tied and low-rank query and key matrices. In the asymptotic limit of high-dimensional data and a comparably large number of training samples we provide a tight closed-form characterization of the global minimum of the non-convex empirical loss landscape. We show that this minimum corresponds to either a positional attention mechanism (with tokens attending to each other based on their respective positions) or a semantic attention mechanism (with tokens attending to each other based on their meaning), and evidence an emergent phase transition from the former to the latter with increasing sample complexity. Finally, we compare the dot-product attention layer to a linear positional baseline, and show that it outperforms the latter using the semantic mechanism provided it has access to sufficient data.

cs.LG↗

arXiv2023

Dynamical Phase Transitions in Graph Cellular Automata

Discrete dynamical systems can exhibit complex behaviour from the iterative application of straightforward local rules. A famous example are cellular automata whose global dynamics are notoriously challenging to analyze. To address this, we relax the regular connectivity grid of cellular automata to a random graph, which gives the class of graph cellular automata. Using the dynamical cavity method (DCM) and its backtracking version (BDCM), we show that this relaxation allows us to derive asymptotically exact analytical results on the global dynamics of these systems on sparse random graphs. Concretely, we showcase the results on a specific subclass of graph cellular automata with ``conforming non-conformist'' update rules, which exhibit dynamics akin to opinion formation. Such rules update a node's state according to the majority within their own neighbourhood. In cases where the majority leads only by a small margin over the minority, nodes may exhibit non-conformist behaviour. Instead of following the majority, they either maintain their own state, switch it, or follow the minority. For configurations with different initial biases towards one state we identify sharp dynamical phase transitions in terms of the convergence speed and attractor types. From the perspective of opinion dynamics this answers when consensus will emerge and when two opinions coexist almost indefinitely.

cond-mat.dis-nn↗

arXiv2023

Unveiling the Hessian's Connection to the Decision Boundary

Understanding the properties of well-generalizing minima is at the heart of deep learning research. On the one hand, the generalization of neural networks has been connected to the decision boundary complexity, which is hard to study in the high-dimensional input space. Conversely, the flatness of a minimum has become a controversial proxy for generalization. In this work, we provide the missing link between the two approaches and show that the Hessian top eigenvectors characterize the decision boundary learned by the neural network. Notably, the number of outliers in the Hessian spectrum is proportional to the complexity of the decision boundary. Based on this finding, we provide a new and straightforward approach to studying the complexity of a high-dimensional decision boundary; show that this connection naturally inspires a new generalization measure; and finally, we develop a novel margin estimation technique which, in combination with the generalization measure, precisely identifies minima with simple wide-margin boundaries. Overall, this analysis establishes the connection between the Hessian and the decision boundary and provides a new method to identify minima with simple wide-margin decision boundaries.

cs.LG↗

arXiv2023

Backtracking Dynamical Cavity Method

The cavity method is one of the cornerstones of the statistical physics of disordered systems such as spin glasses and other complex systems. It is able to analytically and asymptotically exactly describe the equilibrium properties of a broad range of models. Exact solutions for dynamical, out-of-equilibrium properties of disordered systems are traditionally much harder to obtain. Even very basic questions such as the limiting energy of a fast quench are so far open. The dynamical cavity method partly fills this gap by considering short trajectories and leveraging the static cavity method. However, being limited to a couple of steps forward from the initialization it typically does not capture dynamical properties related to attractors of the dynamics. We introduce the backtracking dynamical cavity method that instead of analysing the trajectory forward from initialization, analyses trajectories that are found by tracking them backward from attractors. We illustrate that this rather elementary twist on the dynamical cavity method leads to new insight into some of the very basic questions about the dynamics of complex disordered systems. This method is as versatile as the cavity method itself and we hence anticipate that our paper will open many avenues for future research of dynamical, out-of-equilibrium, properties in complex systems.

cond-mat.dis-nn↗

arXiv2022

(Dis)assortative Partitions on Random Regular Graphs

We study the problem of assortative and disassortative partitions on random $d$-regular graphs. Nodes in the graph are partitioned into two non-empty groups. In the assortative partition every node requires at least $H$ of their neighbors to be in their own group. In the disassortative partition they require less than $H$ neighbors to be in their own group. Using the cavity method based on analysis of the Belief Propagation algorithm we establish for which combinations of parameters $(d,H)$ these partitions exist with high probability and for which they do not. For $H>\lceil \frac{d}{2} \rceil $ we establish that the structure of solutions to the assortative partition problems corresponds to the so-called frozen-1RSB. This entails a conjecture of algorithmic hardness of finding these partitions efficiently. For $H \le \lceil \frac{d}{2} \rceil $ we argue that the assortative partition problem is algorithmically easy on average for all $d$. Further we provide arguments about asymptotic equivalence between the assortative partition problem and the disassortative one, going trough a close relation to the problem of single-spin-flip-stable states in spin glasses. In the context of spin glasses, our results on algorithmic hardness imply a conjecture that gapped single spin flip stable states are hard to find which may be a universal reason behind the observation that physical dynamics in glassy systems display convergence to marginal stability.

cond-mat.dis-nn↗

arXiv2021

Bandits for Learning to Explain from Explanations

We introduce Explearn, an online algorithm that learns to jointly output predictions and explanations for those predictions. Explearn leverages Gaussian Processes (GP)-based contextual bandits. This brings two key benefits. First, GPs naturally capture different kinds of explanations and enable the system designer to control how explanations generalize across the space by virtue of choosing a suitable kernel. Second, Explearn builds on recent results in contextual bandits which guarantee convergence with high probability. Our initial experiments hint at the promise of the approach.

cs.LG↗

arXiv2020

Neurally Augmented ALISTA

It is well-established that many iterative sparse reconstruction algorithms can be unrolled to yield a learnable neural network for improved empirical performance. A prime example is learned ISTA (LISTA) where weights, step sizes and thresholds are learned from training data. Recently, Analytic LISTA (ALISTA) has been introduced, combining the strong empirical performance of a fully learned approach like LISTA, while retaining theoretical guarantees of classical compressed sensing algorithms and significantly reducing the number of parameters to learn. However, these parameters are trained to work in expectation, often leading to suboptimal reconstruction of individual targets. In this work we therefore introduce Neurally Augmented ALISTA, in which an LSTM network is used to compute step sizes and thresholds individually for each target vector during reconstruction. This adaptive approach is theoretically motivated by revisiting the recovery guarantees of ALISTA. We show that our approach further improves empirical performance in sparse reconstruction, in particular outperforming existing algorithms by an increasing margin as the compression ratio becomes more challenging.

cs.LG↗