This work presents a method for end-to-end sequence learning, and more specifically in the framework of Connectionist Temporal Classification (CTC). The paper has two main contributions: - The first is a regularization of the training of the CTC objective in order to reduce the over-confidence of the model. In order to do that, the authors propose a method based on conditional entropy. More specifically, the proposed regularization would encourages the model to explore paths that are close to the dominant one. In order to do so, they suppose that the consecutive elements of a sequence have equal spacing.

The well known maximum-entropy principle due to Jaynes, which states that given mean parameters, the maximum entropy distribution matching them is in an exponential family has been very popular in machine learning due to its "Occam's razor" interpretation. Unfortunately, calculating the potentials in the maximumentropy distribution is intractable [BGS14]. We provide computationally efficient versions of this principle when the mean parameters are pairwise moments: we design distributions that approximately match given pairwise moments, while having entropy which is comparable to the maximum entropy distribution matching those moments. We additionally provide surprising applications of the approximate maximum entropy principle to designing provable variational methods for partition function calculations for Ising models without any assumptions on the potentials of the model. More precisely, we show that we can get approximation guarantees for the log-partition function comparable to those in the low-temperature limit, which is the setting of optimization of quadratic forms over the hypercube.

This report presents a solution for the swing-up and stabilisation tasks of the acrobot and the pendubot, developed for the AI Olympics competition at IROS 2024. Our approach employs the Average-Reward Entropy Advantage Policy Optimization (AR-EAPO), a model-free reinforcement learning (RL) algorithm that combines average-reward RL and maximum entropy RL. Results demonstrate that our controller achieves improved performance and robustness scores compared to established baseline methods in both the acrobot and pendubot scenarios, without the need for a heavily engineered reward function or system model. The current results are applicable exclusively to the simulation stage setup.

Transformers suffer from the computational overhead of their quadratic dependence on the length of sequences processed. We present three methods, all adding an intermediate step between training and inference/generation, which extend the autoregressive length of transformers. All rely on a Maximum Entropy Principle (MEP) whereby entropy is maximized in the presence of suitable constraints, accounted for by use of Lagrange Multipliers. These constraint methods extend the autoregressive character from T to 2T tokens in a linear-with-T fashion. There is overhead associated with this added step, but they should still be faster than the standard methods.

Entropy Regularisation is a widely adopted technique that enhances policy optimisation performance and stability. A notable form of entropy regularisation is augmenting the objective with an entropy term, thereby simultaneously optimising the expected return and the entropy. This framework, known as maximum entropy reinforcement learning (MaxEnt RL), has shown theoretical and empirical successes. However, its practical application in straightforward on-policy actor-critic settings remains surprisingly underexplored. We hypothesise that this is due to the difficulty of managing the entropy reward in practice. This paper proposes a simple method of separating the entropy objective from the MaxEnt RL objective, which facilitates the implementation of MaxEnt RL in on-policy settings. Our empirical evaluations demonstrate that extending Proximal Policy Optimisation (PPO) and Trust Region Policy Optimisation (TRPO) within the MaxEnt framework improves policy optimisation performance in both MuJoCo and Procgen tasks. Additionally, our results highlight MaxEnt RL's capacity to enhance generalisation.

We revisit the classical problem of inverting dimension-reducing linear mappings using the maximum entropy (MaxEnt) criterion. In the literature, solutions are problem-dependent, inconsistent, and use different entropy measures. We propose a new unified approach that not only specializes to the existing approaches, but offers solutions to new cases, such as when data values are constrained to [0, 1], which has new applications in machine learning.

Maximum entropy inference is a widely used method in machine learning, particularly in the context of graphical models (McCallum et al., 2000; Kindermann & Snell, 1980; Ackley et al., 1985; Bresler, 2015; Hamilton et al., 2017) and natural language processing (Berger et al., 1996). In graphical models, it is known as the backward mapping, the problem of computing the model parameters from the marginal information (Wainwright & Jordan, 2007). The inverse problem of estimating marginal parameters from the model parameters is called the forward mapping. Maximum entropy inference is also a core concept in statistical physics (Jaynes, 1957) known as the Jaynes' principle which links statistical mechanics and information theory. The Hammersley-Clifford theorem establishes that, in the classical case, any positive probability distribution satisfying the local Markov property can be represented as a Gibbs distribution (Lafferty et al., 2001).

We present a maximum entropy inverse reinforcement learning (IRL) approach for improving the sample quality of diffusion generative models, especially when the number of generation time steps is small. Similar to how IRL trains a policy based on the reward function learned from expert demonstrations, we train (or fine-tune) a diffusion model using the log probability density estimated from training data. Since we employ an energy-based model (EBM) to represent the log density, our approach boils down to the joint training of a diffusion model and an EBM. Our IRL formulation, named Diffusion by Maximum Entropy IRL (DxMI), is a minimax problem that reaches equilibrium when both models converge to the data distribution. The entropy maximization plays a key role in DxMI, facilitating the exploration of the diffusion model and ensuring the convergence of the EBM. We also propose Diffusion by Dynamic Programming (DxDP), a novel reinforcement learning algorithm for diffusion models, as a subroutine in DxMI. DxDP makes the diffusion model update in DxMI efficient by transforming the original problem into an optimal control formulation where value functions replace back-propagation in time. Our empirical studies show that diffusion models fine-tuned using DxMI can generate high-quality samples in as few as 4 and 10 steps. Additionally, DxMI enables the training of an EBM without MCMC, stabilizing EBM training dynamics and enhancing anomaly detection performance.

Multi-agent reinforcement learning (MARL) tasks often utilize a centralized training with decentralized execution (CTDE) framework. QMIX is a successful CTDE method that learns a credit assignment function to derive local value functions from a global value function, defining a deterministic local policy. However, QMIX is hindered by its poor exploration strategy. While maximum entropy reinforcement learning (RL) promotes better exploration through stochastic policies, QMIX's process of credit assignment conflicts with the maximum entropy objective and the decentralized execution requirement, making it unsuitable for maximum entropy RL. In this paper, we propose an enhancement to QMIX by incorporating an additional local Q-value learning method within the maximum entropy RL framework. Our approach constrains the local Q-value estimates to maintain the correct ordering of all actions. Due to the monotonicity of the QMIX value function, these updates ensure that locally optimal actions align with globally optimal actions. We theoretically prove the monotonic improvement and convergence of our method to an optimal solution. Experimentally, we validate our algorithm in matrix games, Multi-Agent Particle Environment and demonstrate state-of-the-art performance in SMAC-v2.

Quantum machine learning has demonstrated significant potential in solving practical problems, particularly in statistics-focused areas such as data science and finance. However, challenges remain in preparing and learning statistical models on a quantum processor due to issues with trainability and interpretability. In this letter, we utilize the maximum entropy principle to design a statistics-informed parameterized quantum circuit (SI-PQC) for efficiently preparing and training of quantum computational statistical models, including arbitrary distributions and their weighted mixtures. The SI-PQC features a static structure with trainable parameters, enabling in-depth optimized circuit compilation, exponential reductions in resource and time consumption, and improved trainability and interpretability for learning quantum states and classical model parameters simultaneously. As an efficient subroutine for preparing and learning in various quantum algorithms, the SI-PQC addresses the input bottleneck and facilitates the injection of prior knowledge.