arXiv:2505. 11635v2 Announce Type: cross Abstract: Many real-world tasks, from associative memory to symbolic reasoning, benefit from discrete, structured representations that standard continuous latent models can struggle to express.
arXiv:2607. 27077v1 Announce Type: new Abstract: Energy-Based Models (EBMs) provide an interpretable framework for generative modeling of scientific data, but poor Markov Chain Monte Carlo mixing often limits their reliability.
Efficient sampling of molecular systems at thermodynamic equilibrium is a hallmark challenge in statistical physics. This challenge has driven the development of Boltzmann Generators (BGs), which allow rapid generation of uncorrelated equilibrium samples by combining a generative model with exact likelihoods and an importance sampling correction.
Energy-Based Models (EBMs) provide an interpretable framework for generative modeling of scientific data, but poor Markov Chain Monte Carlo mixing often limits their reliability. We introduce a training algorithm based on Parallel Trajectory Tempering (PTT), which exploits the continuity of the optimization path to maintain equilibrium sampling throughout learning.
arXiv:2606. 27361v1 Announce Type: cross Abstract: Efficient sampling of molecular systems at thermodynamic equilibrium is a hallmark challenge in statistical physics.
The paper introduces a new transition kernel for Restricted Boltzmann Machines that operates over the sequence of models used in Deep Tempering. This kernel employs a round‑trip structure, allowing nonlocal moves in a single transition while keeping the RBM sequence unchanged. Experiments demonstrate that it achieves higher sampling quality with fewer transitions than both blocked Gibbs sampling and Deep Tempering, and it stabilizes learning by reducing training failures.
The paper introduces Langevin simulated bifurcation (LSB), a fast, parallel Boltzmann sampler that matches the accuracy of sequential MCMC methods. It also proposes conditional expectation matching (CEM), an efficient technique for estimating the effective temperature of samples from energy‑based models with conditional independence. Building on these, the authors develop sampler adaptive learning (SAL), which adjusts the model temperature to align with the distribution produced by LSB, enabling efficient training of semi‑restricted Boltzmann machines (SRBMs) and outperforming conventional methods on synthetic spin‑glass datasets.
arXiv:2602. 00415v2 Announce Type: replace Abstract: Memory is not merely a storage mechanism for intelligent systems, but a structure for organizing evidence and constraining belief.
The study explores whether a Deep Belief Network (DBN) and a Bidirectional Gated Recurrent Unit (Bi‑GRU) can distinguish four distinct trajectory types in the three‑state majority‑vote model (MV3): approach from disorder, approach from order, departure to disorder, and departure to order. The DBN, trained unsupervised on static equilibrium snapshots, partially separates these trajectories in its 81‑dimensional latent space, while a two‑layer Bi‑GRU trained on sequences of DBN‑encoded snapshots achieves near‑perfect classification, as confirmed by t‑SNE visualizations on both training and test data. A sliding‑window application of the Bi‑GRU to continuous MV3 dynamics further demonstrates real‑time detection of the system’s current dynamical regime.
arXiv:2607. 15682v1 Announce Type: new Abstract: Sampling from an unnormalized Boltzmann density requires proposals that move probability mass globally while retaining enough path-probability information for statistical correction.
arXiv:2607. 10128v1 Announce Type: new Abstract: Recursive reasoning models address structured problems by repeatedly updating latent states of small neural networks.
arXiv:2601. 21026v2 Announce Type: replace-cross Abstract: Sampling configurations at thermodynamic equilibrium is a central challenge in statistical physics.
arXiv:2607. 29503v1 Announce Type: new Abstract: While neural networks are typically evaluated by their training and test performance, these metrics do not reveal how robust a learned representation is.