We study reinforcement learning (RL) with linear function approximation. Existing algorithms for this problem only have high-probability regret and/or Probably Approximately Correct (PAC) sample complexity guarantees, which cannot guarantee the convergence to the optimal policy. In this paper, in order to overcome the limitation of existing algorithms, we propose a new algorithm called FLUTE, which enjoys uniform-PAC convergence to the optimal policy with high probability. The uniform-PAC guarantee is the strongest possible guarantee for reinforcement learning in the literature, which can directly imply both PAC and high probability regret bounds, making our algorithm superior to all existing algorithms with linear function approximation. At the core of our algorithm is a novel minimax value function estimator and a multi-level partition scheme to select the training samples from historical observations. Both of these techniques are new and of independent interest.

We study reinforcement learning (RL) with linear function approximation. Existing algorithms for this problem only have high-probability regret and/or Probably Approximately Correct (PAC) sample complexity guarantees, which cannot guarantee the convergence to the optimal policy. In this paper, in order to overcome the limitation of existing algorithms, we propose a new algorithm called FLUTE, which enjoys uniform-PAC convergence to the optimal policy with high probability. The uniform-PAC guarantee is the strongest possible guarantee for reinforcement learning in the literature, which can directly imply both PAC and high probability regret bounds, making our algorithm superior to all existing algorithms with linear function approximation. At the core of our algorithm is a novel minimax value function estimator and a multi-level partition scheme to select the training samples from historical observations. Both of these techniques are new and of independent interest.

With its superior rendering performance and high-fidelity rendering quality, 3DGS is excelling at its previous NeRF counterparts. The most recent 3DGS method focuses either on improving the instability of rendering efficiency or reducing the model size. On the other hand, the training efficiency of 3DGS on large-scale scenes has not gained much attention. In this work, we propose DOGS, a method that trains 3DGS distributedly. Our method first decomposes a scene into K blocks and then introduces the Alternating Direction Method of Multipliers (ADMM) into the training procedure of 3DGS. During training, our DOGS maintains one global 3DGS model on the master node and K local 3DGS models on the slave nodes. The K local 3DGS models are dropped after training and we only query the global 3DGS model during inference. The training time is reduced by scene decomposition, and the training convergence and stability are guaranteed through the consensus on the shared 3D Gaussians. Our method accelerates the training of 3DGS by 6+ times when evaluated on large-scale scenes while concurrently achieving state-of-the-art rendering quality.

The quantum many-body problem lies at the center of the most important open challenges in condensed matter, quantum chemistry, atomic, nuclear, and highenergy physics. While quantum Monte Carlo, when applicable, remains the most powerful numerical technique capable of treating dozens or hundreds of degrees of freedom with high accuracy, it is restricted to models that are not afflicted by the infamous sign problem. A powerful alternative that has emerged in recent years is the use of neural networks as variational estimators for quantum states. In this work, we propose a symmetry-projected variational solution in the form of linear combinations of simple restricted Boltzmann machines. This construction allows one to explore states outside of the original variational manifold and increase the representation power with moderate computational effort. Besides allowing one to restore spatial symmetries, an expansion in terms of Krylov states using a Lanczos recursion offers a solution that can further improve the quantum state accuracy.

Recent advancements in large language models (LLMs) focus on aligning to heterogeneous human expectations and values via multi-objective preference alignment. However, existing methods are dependent on the policy model parameters, which require high-cost repetition of their alignment algorithms for each new policy model, and they cannot expand to unseen objectives due to their static alignment objectives. In this work, we propose Meta-Objective Aligner (MetaAligner), the first policy-agnostic and generalizable method for multi-objective preference alignment. MetaAligner models multi-objective alignment into three stages: (1) dynamic objectives reformulation algorithm reorganizes traditional alignment datasets to supervise the model on performing flexible alignment across different objectives; (2) conditional weak-to-strong correction paradigm aligns the weak outputs of fixed policy models to approach strong outputs with higher preferences in the corresponding alignment objectives, enabling plug-and-play inferences on any policy models, which significantly reduces training costs and facilitates alignment on close-source policy models; (3) generalizable inference method flexibly adjusts target objectives by updating their text descriptions in the prompts, facilitating generalizable alignment to unseen objectives. Experimental results show that MetaAligner achieves significant and balanced improvements in multiobjective alignments on 10 state-of-the-art policy models, and saves up to 93.63% of GPU training hours compared to previous alignment methods.

This work concerns self-supervised video representation learning (SSVRL), one topic that has received much attention recently. Since videos are storage-intensive and contain a rich source of visual content, models designed for SSVRL are expected to be storage-and computation-efficient, as well as effective. However, most existing methods only focus on one of the two objectives, failing to consider both at the same time. In this work, for the first time, the seemingly contradictory goals are simultaneously achieved by exploiting compressed videos and capturing mutual information between two input streams. Specifically, a novel Motion Vector based Cross Guidance Contrastive learning approach (MVCGC) is proposed. For storage and computation efficiency, we choose to directly decode RGB frames and motion vectors (that resemble low-resolution optical flows) from compressed videos on-the-fly. To enhance the representation ability of the motion vectors, hence the effectiveness of our method, we design a cross guidance contrastive learning algorithm based on multi-instance InfoNCE loss, where motion vectors can take supervision signals from RGB frames and vice versa. Comprehensive experiments on two downstream tasks show that our MVCGC yields new state-of-the-art while being significantly more efficient than its competitors.

We propose a novel type of balanced clustering algorithm to approximate attention.

New Experiments: We thank the reviewers for their valuable comments! Unclear if the proposed method is better than only using LSH. Thank you for the suggestions. ALSH significantly outperforms the E2LSH and the Reformer LSH scheme. SMYRF-BERT base (see also Table 2).

The generative model for our observations is defined as s = Ax + ε. We therefore set β to K φ, where φ is the average magnitude of the kernel functions. For all of our simulations, the kernels were a set of 20 Gaussians with σ = 0.06, centered such that they evenly tile the range from 1 to 1, and β = 9.45. Briefly, this framework proposes a way of embedding a linear dynamical system defined by (3) in the spiking activity of a network of N neurons. The network's estimate of the desired dynamics, ĉ Specifically, each neuron's conditional intensity function is a sigmoidal function of its membrane potential: λ Figure 1 shows a simple demonstration of the decision circuit using samples from a 2D posterior generated by the inference circuit. In this case, the inference circuit is sampling from the posterior of a linear Gaussian model.