A noisy top-k gated mixture-of-experts layer between LSTMs scales neural networks to 137B parameters with sub-linear compute, beating SOTA on language modeling and machine translation.
hub
One billion word benchmark for measuring progress in statistical language modeling
20 Pith papers cite this work. Polarity classification is still indexing.
20
Pith papers citing it
hub tools
representative citing papers
Infinite Mask Diffusion Models use stochastic infinite-state masks to overcome the factorization error lower bound in standard masked diffusion, achieving superior few-step performance on language tasks via distillation.
Spherical vMF flows reduce the continuity equation on the sphere to a scalar ODE in cosine similarity, enabling posterior-weighted sampling of categorical sequences via cross-entropy trained posteriors.
LangFlow is the first continuous diffusion language model to rival discrete diffusion on perplexity and generative perplexity while exceeding autoregressive baselines on several zero-shot tasks.
Lesioning a shared core in multilingual LLMs drops whole-brain fMRI encoding correlation by 60.32%, while language-specific lesions selectively weaken predictions only for the matched native language.
OPT releases open decoder-only transformers up to 175B parameters that match GPT-3 performance at one-seventh the carbon cost, along with code and training logs.
Multi-query attention shares keys and values across heads in Transformers, greatly reducing memory bandwidth for faster decoding with only minor quality loss.
GLUE is a multi-task benchmark for general natural language understanding that includes a diagnostic test suite and finds limited gains from current multi-task learning methods over single-task training.
Deep learning generalization error follows power-law scaling with training set size across multiple domains, with model size scaling sublinearly with data size.
Pointer sentinel-LSTM mixes context copying with softmax prediction to reach 70.9 perplexity on Penn Treebank using fewer parameters than standard LSTMs.
Summing outputs from separately trained QLoRA PEFT modules provides strong performance for attribute-controlled text generation, often matching or exceeding single-task modules even on single-attribute tests.
Joint training of the latent space with the diffusion process produces a competitive latent diffusion language model that is faster than existing discrete and continuous diffusion baselines.
Uniform-based discrete diffusion models behave as associative memories that retrieve unseen data, with a dataset-size-driven memorization-to-generalization transition detectable via conditional entropy of token predictions.
IDDM interpolates diffusion transitions with a resampling mechanism to lessen dependence on intermediate latents and improve sample quality over masked and uniform discrete diffusion models.
Generative perplexity and entropy are shown to be the two additive components of KL divergence to a reference distribution, motivating generative frontiers as a principled evaluation method for diffusion language models.
Properly filtered web data from CommonCrawl alone trains LLMs that significantly outperform models trained on The Pile, with 600 billion tokens and 1.3B/7.5B parameter models released.
Language models show good calibration when asked to estimate the probability that their own answers are correct, with performance improving as models get larger.
Ranked preference modeling outperforms imitation learning for language model alignment and scales more favorably with model size.
Re-evaluating controlled text generation systems under standardized conditions reveals that many published performance claims do not hold, highlighting the need for consistent evaluation practices.
FastOmniTMAE parallelizes clause learning in Tsetlin Machine autoencoders to achieve up to 5x faster training with comparable embedding quality and low-footprint FPGA deployment.
citing papers explorer
-
Outrageously Large Neural Networks: The Sparsely-Gated Mixture-of-Experts Layer
A noisy top-k gated mixture-of-experts layer between LSTMs scales neural networks to 137B parameters with sub-linear compute, beating SOTA on language modeling and machine translation.
-
Deep Learning Scaling is Predictable, Empirically
Deep learning generalization error follows power-law scaling with training set size across multiple domains, with model size scaling sublinearly with data size.