AURORA-LM is introduced, a continuous-latent diffusion language model that separates the construction of a decodable text representation from the modeling of its distribution, and achieves the strongest performance among evaluated continuous and diffusion-based language models on OpenWebText free generation and XSum summarization.
Abstract
Language remains an outlier in generative modeling: while images, video, and audio are increasingly modeled in continuous latent spaces, text generation still relies predominantly on discrete tokens. Existing continuous language models either inherit embedding spaces not designed for joint generation and decoding, or compress autoencoded latents to ease diffusion, sacrificing token-level fidelity. Instead of simplifying the representation to suit the generative model, we preserve a high-capacity, decodable text latent and design the diffusion model to learn its distribution directly. We introduce AURORA-LM, a continuous-latent diffusion language model that separates the construction of a decodable text representation from the modeling of its distribution. A Query-based Encoder-Decoder organizes text into a high-capacity, prefix-aligned latent sequence, and a Block-causal Diffusion Transformer learns its distribution through flow matching, generating blocks left to right while denoising positions within each block in parallel. Because such a latent is harder for diffusion to model, AURORA-LM restricts only the noisy-input pathway while retaining the full clean-latent prediction target, accommodating full-width latents without reducing decoder-facing capacity. We further calibrate the noise-level distribution to the latent width, and introduce self-trajectory consistency to bridge independently sampled training noise and iterative denoising at inference. AURORA-LM achieves the strongest performance among evaluated continuous and diffusion-based language models on OpenWebText free generation and XSum summarization. Scaling to 1B parameters with about 1500 EFLOPs of total compute yields further gains, surpassing a larger publicly released latent-diffusion language model under a matched evaluation protocol. All experiments are conducted on Ascend NPUs.
V-RAE, a video representation autoencoder that builds compact generative latents on top of frozen vision foundation model representations, and tFVD, a temporal-coherence diagnostic that correlates more reliably with downstream generation quality are introduced.
Latent diffusion modeling (LDM), a prominent paradigm, utilizes tokenizers to map input signal to compressed representation. This dependency positions tokenizer as an integral part of generation process itself, since it affects learning speed, quality of synthesized samples and lay foundation for later applications. This report presents series of KVAE tokenizers for audio, image and video, all designed for subsequent text-conditioned generation: KVAE-Audio, a continuous full-band 48 kHz tokenizer with a 50 Hz latent of 64 channels; KVAE-3D -- two causal video tokenizers for 4x16x16 and 4x8x8 compression; KVAE-2D, an image model, compressing input by factor of 8 with 32 channels. We demonstrate that reconstruction (PSNR, LPIPS, PESQ, etc.) and generation results on objective (Frechet Distance, CLIP score, CLAP score, etc.) and subjective (side-by-side evaluation) metrics matches or surpasses frontier opensource tokenizers, such as VAEs from Wan-2.2, HunyuanVideo-1.5, FLUX.2, MovieGen, StableAudio and MMAudio. Considering difficulty of development, we share with community training details, model selection method and ablation on design choices. The code is publicly available at https://github.com/kandinskylab/kvae and https://github.com/kandinskylab/kvae-audio.
Andrey Shutkin, Denis Parkhomenko, Ivan Kirillov et al.· 0 citations
Latent diffusion models have emerged as a dominant framework for high-fidelity image and video synthesis, operating in compact latent spaces with variational autoencoders (VAEs) to enhance computational efficiency without compromising visual quality. However, conventional VAEs are suboptimal for video data as they employ fixed compression ratios that cannot adapt to the varying complexity of spatio-temporal content. We present KATok (Keep-or-Drop? Adaptive Tokenizer for Compact Video Representation), a transformer-based VAE that incorporates an adaptive token selector which is jointly learned with latent tokens. By evaluating each token's content-richness as keep-or-drop probability, the token selector effectively discards uninformative tokens, naturally allowing data-dependent compression. Applying adaptive tokenization to diffusion models may cause spatial misalignment, as token dropping can disturb the original spatio-temporal structure. To alleviate this issue, we propose two position-prediction strategies: cascaded and joint generation, to ensure spatial consistency. We empirically show that our model achieves strong reconstruction and generation quality at a state-of-the-art compression ratio. Further analysis on video data reveals that this improvement is primarily achieved by reducing spatio-temporal redundancy and removing uninformative tokens, as supported by both quantitative and qualitative results.
Yeonkyeong Lee, Hyun-Young Go, Jongmin Kim et al.· 0 citations
Every audio generative system makes two coupled decisions: what representation to generate, and how to model its distribution. This paper organizes audio generative modeling around this coupling. For representation design, we compare discrete, continuous, and hybrid latents through four objectives: representation burden, distortion, empirical modelability, and streaming compatibility. For distribution modeling, rather than treating a latent's difficulty as an intrinsic scalar, we use two diagnostic dimensions: dependency horizon, how far useful context extends, and conditional ambiguity, how much uncertainty remains after conditioning. These dimensions refine the common semantic-versus-acoustic intuition: variables with a long dependency horizon should receive global modeling capacity. Conditionally ambiguous detail may be delegated to a local or iterative generator. Applied to representative systems, this view shows that RVQ's residual order gives ordered capacity but not ordered semantics, that AudioLM's semantic-versus-acoustic cascade is one explicit placement of this boundary rather than a universal template, and that autoregression, iterative refinement, and hybrid designs differ chiefly in how they trade dependency horizon against critical-path generation cost. The distinction between discrete and continuous latents describes the output interface; dependency horizon, conditional ambiguity, and streaming determine how that interface should be modeled. Rather than cataloguing individual systems, we provide an evaluation and design framework for comparing representation-model pairs.
Text-to-motion generation aims to synthesize semantically consistent and naturally coherent motion sequences from natural language descriptions. Given the continuous nature of human motion, diffusion models operating in a continuous latent space offer inherent advantages over vector quantization-based methods, particularly in avoiding quantization errors and in modeling quality. However, existing diffusion models primarily rely on mean squared error loss. This stepwise regression paradigm often leads to ‘over-smoothed’ motion sequences and struggles to capture the subtle semantic nuances embedded in textual descriptions. To realize the potential for continuous diffusion generation, an enhanced latent-space diffusion framework designed to elevate generation capabilities across two dimensions, namely, distribution approximation and semantic alignment, is proposed. Specifically, a latent-space adversarial discriminator is incorporated. By applying decoupled adversarial supervision, this component mitigates the detail loss caused by mean regression, significantly enhancing the physical realism and dynamic sharpness. Concurrently, a latent-space contrastive alignment strategy is introduced during the denoising process that reinforces the correspondence of the generated motion sequences with the given textual inputs via explicit cross-modal constraints. Extensive experiments on standard benchmarks demonstrate that the proposed method effectively addresses the limitations of conventional diffusion models, thus validating the potential of continuous diffusion frameworks within the domain of text-driven motion synthesis.
Zhaowu Li, Rui Liu, Deheng Zhu et al.· Visual Computing for Industr...· 0 citations
A novel framework that performs selective cross-modal alignment through a learnable masking mechanism, enabling the model to isolate and align only the shared latent components relevant to both modalities is proposed.
Jiyang Zheng, Siqi Pan, Yu Yao et al.· Neural Information Processin...· 6 citations
