ICML 2024 Papers

Zero-shot quantization aims to learn a quantized model from a pre-trained full-precision model with no access to original real training data. The common idea in zero-shot quantization approaches is to generate synthetic data for quantizing the full-precision model. While it is well-known that deep neural networks with low sharpness have better generalization ability, none of the previous zero-shot quantization works considers the sharpness of the quantized model as a criterion for generating training data. This paper introduces a novel methodology that takes into account quantized model sharpness in synthetic data generation to enhance generalization. Specifically, we first demonstrate that sharpness minimization can be attained by maximizing gradient matching between the reconstruction loss gradients computed on synthetic and real validation data, under certain assumptions. We then circumvent the problem of the gradient matching without real validation set by approximating it with the gradient matching between each generated sample and its neighbors. Experimental evaluations on CIFAR-100 and ImageNet datasets demonstrate the superiority of the proposed method over the state-of-the-art techniques in low-bit quantization settings.

We introduce HyperFields, a method for generating text-conditioned Neural Radiance Fields (NeRFs) with a single forward pass and (optionally) some fine-tuning. Key to our approach are: (i) a dynamic hypernetwork, which learns a smooth mapping from text token embeddings to the space of NeRFs; (ii) NeRF distillation training, which distills scenes encoded in individual NeRFs into one dynamic hypernetwork. These techniques enable a single network to fit over a hundred unique scenes. We further demonstrate that HyperFields learns a more general map between text and NeRFs, and consequently is capable of predicting novel in-distribution and out-of-distribution scenes --- either zero-shot or with a few finetuning steps. Finetuning HyperFields benefits from accelerated convergence thanks to the learned general map, and is capable of synthesizing novel scenes 5 to 10 times faster than existing neural optimization-based methods. Our ablation experiments show that both the dynamic architecture and NeRF distillation are critical to the expressivity of HyperFields.

Deep neural networks are applied in more and more areas of everyday life. However, they still lack essential abilities, such as robustly dealing with spatially transformed input signals. Approaches to mitigate this severe robustness issue are limited to two pathways: Either models are implicitly regularised by increased sample variability (data augmentation) or explicitly constrained by hard-coded inductive biases. The limiting factor of the former is the size of the data space, which renders sufficient sample coverage intractable. The latter is limited by the engineering effort required to develop such inductive biases for every possible scenario. Instead, we take inspiration from human behaviour, where percepts are modified by mental or physical actions during inference. We propose a novel technique to emulate such an inference process for neural nets. This is achieved by traversing a sparsified inverse transformation tree during inference using parallel energy-based evaluations. Our proposed inference algorithm, called Inverse Transformation Search (ITS), is model-agnostic and equips the model with zero-shot pseudo-invariance to spatially transformed inputs. We evaluated our method on several benchmark datasets, including a synthesised ImageNet test set. ITS outperforms the utilised baselines on all zero-shot test scenarios.

A main theoretical puzzle is why over-parameterized Neural Networks (NNs) generalize well when trained to zero loss (i.e., so they interpolate the data). Usually, the NN is trained with Stochastic Gradient Descent (SGD) or one of its variants. However, recent empirical work examined the generalization of a random NN that interpolates the data: the NN was sampled from a seemingly uniform prior over the parameters, conditioned on that the NN perfectly classifying the training set. Interestingly, such a NN sample typically generalized as well as SGD-trained NNs. We prove that such a random NN interpolator typically generalizes well if there exists an underlying narrow ``teacher NN" that agrees with the labels. Specifically, we show that such a `flat' prior over the NN parametrization induces a rich prior over the NN functions, due to the redundancy in the NN structure. In particular, this creates a bias towards simpler functions, which require less relevant parameters to represent --- enabling learning with a sample complexity approximately proportional to the complexity of the teacher (roughly, the number of non-redundant parameters), rather than the student's.

Despite the remarkable capabilities, Large Language Models (LLMs) face deployment challenges due to their extensive size. Pruning methods drop a subset of weights to accelerate, but many of them require retraining, which is prohibitively expensive and computationally demanding. Recently, post-training pruning approaches introduced novel metrics, enabling the pruning of LLMs without retraining. However, these metrics require the involvement of human experts and tedious trial and error. To efficiently identify superior pruning metrics, we develop an automatic framework for searching symbolic pruning metrics using genetic programming. In particular, we devise an elaborate search space encompassing the existing pruning metrics to discover the potential symbolic pruning metric. We propose an opposing operation simplification strategy to increase the diversity of the population. In this way, Pruner-Zero allows auto-generation of symbolic pruning metrics. Based on the searched results, we explore the correlation between pruning metrics and performance after pruning and summarize some principles. Extensive experiments on LLaMA and LLaMA-2 on language modeling and zero-shot tasks demonstrate that our Pruner-Zero obtains superior performance than SOTA post-training pruning methods. Code at: https://github.com/pprp/Pruner-Zero.

Denoising Diffusion Probabilistic Models (DDPMs) exhibit remarkable capabilities in image generation, with studies suggesting that they can generalize by composing latent factors learned from the training data. In this work, we go further and study DDPMs trained on strictly separate subsets of the data distribution with large gaps on the support of the latent factors. We show that such a model can effectively generate images in the unexplored, intermediate regions of the distribution. For instance, when trained on clearly smiling and non-smiling faces, we demonstrate a sampling procedure which can generate slightly smiling faces without reference images (zero-shot interpolation). We replicate these findings for other attributes as well as other datasets. [Our code is available on GitHub](https://github.com/jdeschena/ddpm-zero-shot-interpolation).

In preference-based Reinforcement Learning (RL), obtaining a large number of preference labels are both time-consuming and costly. Furthermore, the queried human preferences cannot be utilized for the new tasks. In this paper, we propose Zero-shot Cross-task Preference Alignment and Robust Reward Learning (PEARL), which learns policies from cross-task preference transfer without any human labels of the target task. Our contributions include two novel components that facilitate the transfer and learning process. The first is Cross-task Preference Alignment (CPA), which transfers the preferences between tasks via optimal transport. The key idea of CPA is to use Gromov-Wasserstein distance to align the trajectories between tasks, and the solved optimal transport matrix serves as the correspondence between trajectories. The target task preferences are computed as the weighted sum of source task preference labels with the correspondence as weights. Moreover, to ensure robust learning from these transferred labels, we introduce Robust Reward Learning (RRL), which considers both reward mean and uncertainty by modeling rewards as Gaussian distributions. Empirical results on robotic manipulation tasks from Meta-World and Robomimic demonstrate that our method is capable of transferring preference labels across tasks accurately and then learns well-behaved policies. Notably, our approach significantly exceeds existing methods when there are few human preferences. The code and videos of our method are available at: https://sites.google.com/view/pearl-preference.

Early time classification algorithms aim to label a stream of features without processing the full input stream, while maintaining accuracy comparable to that achieved by applying the classifier to the entire input. In this paper, we introduce a statistical framework that can be applied to any sequential classifier, formulating a calibrated stopping rule. This data-driven rule attains finite-sample, distribution-free control of the accuracy gap between full and early-time classification. We start by presenting a novel method that builds on the Learn-then-Test calibration framework to control this gap marginally, on average over i.i.d. instances. As this algorithm tends to yield an excessively high accuracy gap for early halt times, our main contribution is the proposal of a framework that controls a stronger notion of error, where the accuracy gap is controlled conditionally on the accumulated halt times. Numerical experiments demonstrate the effectiveness, applicability, and usefulness of our method. We show that our proposed early stopping mechanism reduces up to 94% of timesteps used for classification while achieving rigorous accuracy gap control.

The accurate classification of quantum states is crucial for advancing quantum computing, as it allows for the effective analysis and correct functioning of quantum devices by analyzing the statistics of the data from quantum measurements. Traditional supervised methods, which rely on extensive labeled measurement outcomes, are used to categorize unknown quantum states with different properties. However, the labeling process demands computational and memory resources that increase exponentially with the number of qubits. We propose SSL4Q, manage to achieve (for the first time) semi-supervised learning specifically designed for quantum state classification. SSL4Q's architecture is tailored to ensure permutation invariance for unordered quantum measurements and maintain robustness in the face of measurement uncertainties. Our empirical studies encompass simulations on two types of quantum systems: the Heisenberg Model and the Variational Quantum Circuit (VQC) Model, with system size reaching up to 50 qubits. The numerical results demonstrate SSL4Q's superiority over traditional supervised models in scenarios with limited labels, highlighting its potential in efficiently classifying quantum states with reduced computational and resource overhead.

Pre-trained vision transformers have revolutionized few-shot image classification, and it has been recently demonstrated that the previous common practice of meta-learning in synergy with these pre-trained transformers still holds significance. In this work, we design a new framework centered exclusively on self-attention, called MetaFormer, which extends the vision transformers beyond patch token interactions to encompass relationships between samples and tasks simultaneously for further advancing their downstream task performance. Leveraging the intrinsical property of ViTs in handling local patch relationships, we propose Masked Sample Attention (MSA) to efficiently embed the sample relationships into the network, where an adaptive mask is attached for enhancing task-specific feature consistency and providing flexibility in switching between few-shot learning setups. To encapsulate task relationships while filtering out background noise, Patch-grained Task Attention (PTA) is designed to maintain a dynamic knowledge pool consolidating diverse patterns from historical tasks. MetaFormer demonstrates coherence and compatibility with off-the-shelf pre-trained vision transformers and shows significant improvements in both inductive and transductive few-shot learning scenarios, outperforming state-of-the-art methods by up to 8.77% and 6.25% on 12 in-domain and 10 cross-domain datasets, respectively.