Neural Discrete Representation Learning
read the original abstract
Learning useful representations without supervision remains a key challenge in machine learning. In this paper, we propose a simple yet powerful generative model that learns such discrete representations. Our model, the Vector Quantised-Variational AutoEncoder (VQ-VAE), differs from VAEs in two key ways: the encoder network outputs discrete, rather than continuous, codes; and the prior is learnt rather than static. In order to learn a discrete latent representation, we incorporate ideas from vector quantisation (VQ). Using the VQ method allows the model to circumvent issues of "posterior collapse" -- where the latents are ignored when they are paired with a powerful autoregressive decoder -- typically observed in the VAE framework. Pairing these representations with an autoregressive prior, the model can generate high quality images, videos, and speech as well as doing high quality speaker conversion and unsupervised learning of phonemes, providing further evidence of the utility of the learnt representations.
This paper has not been read by Pith yet.
Forward citations
Cited by 28 Pith papers
-
ENSEMBITS: an alphabet of protein conformational ensembles
Ensembits creates a discrete vocabulary for protein conformational ensembles that outperforms static tokenizers on dynamics prediction tasks and enables ensemble token prediction from single structures via distillation.
-
ENSEMBITS: an alphabet of protein conformational ensembles
Ensembits is the first tokenizer of protein conformational ensembles that outperforms static tokenizers on RMSF prediction and matches them on function and mutation tasks while using less pretraining data.
-
Masked-Token Prediction for Anomaly Detection at the Large Hadron Collider
The work demonstrates masked-token prediction with transformers for model-independent anomaly detection in LHC data, achieving strong results on top-rich BSM signatures like four-top production using VQ-VAE tokenization.
-
Neuro-Symbolic ODE Discovery with Latent Grammar Flow
Latent Grammar Flow discovers ODEs by placing grammar-based equation representations in a discrete latent space, using a behavioral loss to cluster similar equations, and sampling via a discrete flow model guided by d...
-
Hierarchical Text-Conditional Image Generation with CLIP Latents
A hierarchical prior-decoder model using CLIP latents generates more diverse text-conditional images than direct methods while preserving photorealism and caption fidelity.
-
GLIDE: Towards Photorealistic Image Generation and Editing with Text-Guided Diffusion Models
A 3.5-billion-parameter diffusion model with classifier-free guidance generates images preferred over DALL-E by human raters and can be fine-tuned for text-guided inpainting.
-
Diffusion Models Beat GANs on Image Synthesis
Diffusion models with architecture improvements and classifier guidance achieve superior FID scores to GANs on unconditional and conditional ImageNet image synthesis.
-
Scaling Laws for Autoregressive Generative Modeling
Autoregressive transformers follow power-law scaling laws for cross-entropy loss with nearly universal exponents relating optimal model size to compute budget across four domains.
-
Network-Efficient World Model Token Streaming
An adaptive delta-prioritization algorithm using cosine distance and Hamming-drift thresholds improves embedding distortion by 4.8-7.2% and next-token perplexity by 2.1-6.3% over periodic keyframing at matched low bit...
-
CASCADE: Context-Aware Relaxation for Speculative Image Decoding
CASCADE formalizes semantic interchangeability and convergence in target model representations to enable context-aware acceptance relaxation in tree-based speculative decoding, delivering up to 3.6x speedup on text-to...
-
ASTRA: Mapping Art-Technology Institutions via Conceptual Axes, Text Embeddings, and Unsupervised Clustering
ASTRA combines an eight-axis conceptual framework with text embeddings and unsupervised clustering to map and group 78 art-technology institutions into coherent thematic clusters.
-
FAST: Efficient Action Tokenization for Vision-Language-Action Models
FAST applies discrete cosine transform to robot action sequences for efficient tokenization, enabling autoregressive VLAs to succeed on high-frequency dexterous tasks and scale to 10k hours of data while matching diff...
-
TD-MPC2: Scalable, Robust World Models for Continuous Control
TD-MPC2 scales an implicit world-model RL method to a 317M-parameter agent that masters 80 tasks across four domains with a single hyperparameter configuration.
-
Shap-E: Generating Conditional 3D Implicit Functions
Shap-E encodes 3D assets into implicit function parameters then uses a conditional diffusion model to generate new ones from text, enabling fast multi-representation 3D asset creation.
-
Is Conditional Generative Modeling all you need for Decision-Making?
Return-conditional diffusion models for policies outperform offline RL on benchmarks by circumventing dynamic programming and enable constraint or skill composition.
-
Latent Video Diffusion Models for High-Fidelity Long Video Generation
Latent-space hierarchical diffusion models with targeted error-correction techniques generate realistic videos exceeding 1000 frames while using less compute than prior pixel-space approaches.
-
Language Models (Mostly) Know What They Know
Language models show good calibration when asked to estimate the probability that their own answers are correct, with performance improving as models get larger.
-
A General Language Assistant as a Laboratory for Alignment
Ranked preference modeling outperforms imitation learning for language model alignment and scales more favorably with model size.
-
Vector-quantized Image Modeling with Improved VQGAN
Improved ViT-VQGAN enables autoregressive Transformer pretraining on ImageNet tokens to reach IS 175.1 and FID 4.17 for generation plus 73.2% linear-probe accuracy, beating prior iGPT models.
-
Scaling Laws for Transfer
Effective data transferred from pre-training to fine-tuning is described by a power law in model parameter count and fine-tuning dataset size, acting like a multiplier on the fine-tuning data.
-
Video as Natural Augmentation: Towards Unified AI-Generated Image and Video Detection
VINA trains a single detector on images plus video frames using a cross-modal supervised contrastive objective, yielding bidirectional gains and SOTA results on 14 image, video, and in-the-wild benchmarks.
-
Second-Order Multi-Level Variance Correction for Modality Competition in Multimodal Models
Introduces ML-FOP-SOAP optimizer using Fisher-Orthogonal Projection and hierarchical folding to mitigate modality competition in multimodal autoregressive training, reporting gains over AdamW on Janus and Emu3.
-
PixelFlowCast: Latent-Free Precipitation Nowcasting via Pixel Mean Flows
PixelFlowCast delivers high-fidelity precipitation nowcasts from radar sequences using a latent-free Pixel Mean Flows predictor guided by a deterministic coarse stage and KANCondNet features.
-
SID-Coord: Coordinating Semantic IDs for ID-based Ranking in Short-Video Search
SID-Coord coordinates semantic IDs with hashed item IDs via attention fusion, adaptive gating, and interest alignment, yielding +0.664% long-play rate and +0.369% playback duration gains in production search ranking.
-
From Video to Control: A Survey of Learning Manipulation Interfaces from Temporal Visual Data
A survey introduces an interface-centric taxonomy for video-to-control methods in robotic manipulation and identifies the robotics integration layer as the central open challenge.
-
PaCo-FR: Patch-Pixel Aligned End-to-End Codebook Learning for Facial Representation Pre-training
PaCo-FR introduces a structured-masking and patch-codebook framework for unsupervised facial representation pre-training that claims state-of-the-art results on multiple facial tasks after training on only 2 million u...
-
LASAR: Towards Spatio-temporal Reasoning with Latent Cognitive Map
LASAR pairs a dual-memory system with spatio-temporal contrastive learning to induce latent cognitive maps, reporting 2-3.5% zero-shot gains on VLN-CE and VSI-Bench plus high map self-consistency.
-
Autoencoding sensory substitution
Deep recurrent autoencoders convert images to shortened audio signals that incorporate hearing models, enabling above-chance hand posture discrimination and object reaching after a few hours of training instead of months.
discussion (0)
Sign in with ORCID, Apple, or X to comment. Anyone can read and Pith papers without signing in.