Related papers: Lossless Acceleration for Seq2seq Generation with …

Conventional neural autoregressive decoding commonly assumes a fixed left-to-right generation order, which may be sub-optimal. In this work, we propose a novel decoding algorithm -- InDIGO -- which supports flexible sequence generation in…

Large language models (LLMs) have become ubiquitous in practice and are widely used for generation tasks such as translation, summarization and instruction following. However, their enormous size and reliance on autoregressive decoding…

Text-to-Motion (T2M) generation aims to synthesize realistic and semantically aligned human motion sequences from natural language descriptions. However, current approaches face dual challenges: Generative models (e.g., diffusion models)…

Autoregressive large language models (LLMs) deliver strong performance but require inherently sequential decoding, leading to high inference latency and poor GPU utilization. Speculative decoding mitigates this bottleneck by using a fast…

Tensor decomposition, a collection of factorization techniques for multidimensional arrays, are among the most general and powerful tools for scientific analysis. However, because of their increasing size, today's data sets require more…

While sequence-to-sequence (seq2seq) models achieve state-of-the-art performance in many natural language processing tasks, they can be too slow for real-time applications. One performance bottleneck is predicting the most likely next token…

Non-autoregressive Transformer is a promising text generation model. However, current non-autoregressive models still fall behind their autoregressive counterparts in translation quality. We attribute this accuracy gap to the lack of…

The recently proposed Sequence-to-Sequence (seq2seq) framework advocates replacing complex data processing pipelines, such as an entire automatic speech recognition system, with a single neural network trained in an end-to-end fashion. In…

Recently, Vector Quantized AutoRegressive (VQ-AR) models have shown remarkable results in text-to-image synthesis by equally predicting discrete image tokens from the top left to bottom right in the latent space. Although the simple…

In this paper, we propose a novel deep generative approach to cross-modal retrieval to learn hash functions in the absence of paired training samples through the cycle consistency loss. Our proposed approach employs adversarial training…

Neural text generation models are typically trained by maximizing log-likelihood with the sequence cross entropy (CE) loss, which encourages an exact token-by-token match between a target sequence with a generated sequence. Such training…

Autoregressive decoding in Large Language Models (LLMs) generates one token per step, causing high inference latency. Speculative decoding (SD) mitigates this through a guess-and-verify strategy, but existing training-free variants face…

Implicit Neural Representations (INRs) offer exceptional fidelity for video compression by learning per-video optimized functions, but their adoption is crippled by impractically slow encoding times. Existing attempts to accelerate INR…

Speculative decoding has become the standard approach for accelerating Large Language Model (LLM) inference. It exploits a lossless draft-then-verify procedure to circumvent the latency of autoregressive decoding, achieving impressive…

In this work, we address the task of unconditional head motion generation to animate still human faces in a low-dimensional semantic space from a single reference pose. Different from traditional audio-conditioned talking head generation…

We present Entropy Adaptive Decoding (EAD), a novel approach for efficient language model inference that dynamically switches between different-sized models based on prediction uncertainty. By monitoring rolling entropy in model logit…

Guessing Random Additive Noise Decoding (GRAND) is a recently proposed universal decoding algorithm for linear error correcting codes. Since GRAND does not depend on the structure of the code, it can be used for any code encountered in…

Existing large language models have to run K times to generate a sequence of K tokens. In this paper, we present RecycleGPT, a generative language model with fast decoding speed by recycling pre-generated model states without running the…

We introduce Autoregressive Retrieval Augmentation (AR-RAG), a novel paradigm that enhances image generation by autoregressively incorporating knearest neighbor retrievals at the patch level. Unlike prior methods that perform a single,…

In real-world applications of large language models, outputs are often required to be confined: selecting items from predefined product or document sets, generating phrases that comply with safety standards, or conforming to specialized…