pith. sign in

arxiv: 2205.05131 · v3 · pith:WHDFRWDAnew · submitted 2022-05-10 · 💻 cs.CL

UL2: Unifying Language Learning Paradigms

classification 💻 cs.CL
keywords pre-trainingmodelsobjectivesmodelachieveacrossdifferentdiverse
0
0 comments X
read the original abstract

Existing pre-trained models are generally geared towards a particular class of problems. To date, there seems to be still no consensus on what the right architecture and pre-training setup should be. This paper presents a unified framework for pre-training models that are universally effective across datasets and setups. We begin by disentangling architectural archetypes with pre-training objectives -- two concepts that are commonly conflated. Next, we present a generalized & unified perspective for self-supervision in NLP and show how different pre-training objectives can be cast as one another and how interpolating between different objectives can be effective. We then propose Mixture-of-Denoisers (MoD), a pre-training objective that combines diverse pre-training paradigms together. We furthermore introduce a notion of mode switching, wherein downstream fine-tuning is associated with specific pre-training schemes. We conduct extensive ablative experiments to compare multiple pre-training objectives and find that our method pushes the Pareto-frontier by outperforming T5 & GPT-like models across multiple diverse setups. By scaling our model up to 20B parameters, we achieve SOTA performance on 50 well-established supervised finetuning based NLP tasks. Our model also achieve strong results at in-context learning, outperforming 175B GPT-3 on zero-shot SuperGLUE and tripling the performance of T5-XXL on one-shot summarization. On 0-shot MMLU, UL2 20B outperforms T0 and T5 models. UL2 20B also works well with chain-of-thought prompting and reasoning, making it an appealing choice for research into reasoning at a small to medium scale of 20B parameters. Finally, we apply FLAN instruction tuning to the UL2 20B model, achieving MMLU and Big-Bench scores competitive to FLAN-PaLM 62B. We release Flax-based T5X checkpoints for the UL2 20B & Flan-UL2 20B.

This paper has not been read by Pith yet.

discussion (0)

Sign in with ORCID, Apple, or X to comment. Anyone can read and Pith papers without signing in.

Forward citations

Cited by 20 Pith papers

Reviewed papers in the Pith corpus that reference this work. Sorted by Pith novelty score.

  1. Chain-of-Thought Prompting Elicits Reasoning in Large Language Models

    cs.CL 2022-01 accept novelty 9.0

    Chain-of-thought prompting, by including intermediate reasoning steps in few-shot examples, elicits strong reasoning abilities in large language models on arithmetic, commonsense, and symbolic tasks.

  2. Large Language Model Selection with Limited Annotations

    cs.CL 2026-05 unverdicted novelty 7.0

    SELECT-LLM is the first active model selection framework for LLMs that uses expected information gain from pairwise output similarities to minimize required annotations, reporting up to 84.78% cost reduction across 23...

  3. Leveraging Pretrained Language Models as Energy Functions for Glauber Dynamics Text Diffusion

    cs.LG 2026-05 unverdicted novelty 7.0

    Pretrained language models are used as energy functions for Glauber dynamics in discrete text diffusion, improving generation quality over prior diffusion LMs and matching autoregressive models on benchmarks and reaso...

  4. Semantic Integrity Matters: Benchmarking and Preserving High-Density Reasoning in KV Cache Compression

    cs.CL 2025-02 unverdicted novelty 7.0

    KV cache compression causes task-dependent degradation in high-density reasoning due to disrupted CoT links; ShotKV mitigates this by preserving few-shot examples as indivisible semantic units through phase separation...

  5. Demystifying Training-Time Augmentation for Data-Constrained Language Model Pretraining

    cs.LG 2026-06 unverdicted novelty 6.0

    Training-time augmentations in token noise, permutation, and offset categories reduce overfitting and improve minimum validation loss in multi-epoch autoregressive pretraining on fixed corpora.

  6. UniMaia: Steering Chess Policies with Language for Human-like Play

    cs.CL 2026-05 unverdicted novelty 6.0

    UniMaia demonstrates feasible prompt-conditioned control of a frozen Lc0 chess policy network via text encoder and conditioning, achieving SOTA expected accuracy on prompt benchmarks while remaining competitive on hum...

  7. S^2tory: Story Spine Distillation for Movie Script Summarization

    cs.CL 2026-05 unverdicted novelty 6.0

    S^2tory uses narratological theory and a Narrative Expert Agent to identify plot nuclei in movie scripts for high-fidelity summarization at 3.5x compression, with strong zero-shot generalization to books.

  8. EmbeddingGemma: Powerful and Lightweight Text Representations

    cs.CL 2025-09 unverdicted novelty 6.0

    A 300M-parameter open embedding model sets new SOTA on MTEB for its size class and matches models twice as large while staying effective when compressed.

  9. Better & Faster Large Language Models via Multi-token Prediction

    cs.CL 2024-04 conditional novelty 6.0

    Multi-token prediction training yields higher sample efficiency, better benchmark scores on code generation, and up to 3x faster inference than standard next-token prediction for LLMs.

  10. Scaling Data-Constrained Language Models

    cs.CL 2023-05 conditional novelty 6.0

    Repeating training data up to 4 epochs yields negligible loss increase versus unique data for fixed compute, and a new scaling law accounts for the decaying value of repeated tokens and excess parameters.

  11. CodeT5+: Open Code Large Language Models for Code Understanding and Generation

    cs.CL 2023-05 conditional novelty 6.0

    CodeT5+ is a flexible encoder-decoder LLM family for code pretrained with diverse objectives on multilingual corpora and initialized from existing LLMs, achieving state-of-the-art results on code generation, completio...

  12. BloombergGPT: A Large Language Model for Finance

    cs.LG 2023-03 conditional novelty 6.0

    BloombergGPT is a 50B parameter LLM trained on a 708B token mixed financial and general dataset that outperforms prior models on financial benchmarks while preserving general LLM performance.

  13. The Flan Collection: Designing Data and Methods for Effective Instruction Tuning

    cs.AI 2023-01 conditional novelty 6.0

    The Flan Collection demonstrates that task balancing, data enrichment, and mixed prompt training are critical to effective instruction tuning, yielding stronger Flan-T5 models released publicly.

  14. Efficient Training of Language Models to Fill in the Middle

    cs.CL 2022-07 unverdicted novelty 6.0

    Autoregressive language models trained on data with middle spans relocated to the end learn infilling without degrading left-to-right perplexity or sampling quality.

  15. Emergent Abilities of Large Language Models

    cs.CL 2022-06 unverdicted novelty 6.0

    Emergent abilities are capabilities present in large language models but absent in smaller ones and cannot be predicted by extrapolating smaller model performance.

  16. Open-Sora: Democratizing Efficient Video Production for All

    cs.CV 2024-12 unverdicted novelty 5.0

    Open-Sora releases an open-source video generation model based on a Spatial-Temporal Diffusion Transformer that decouples spatial and temporal attention, supporting text-to-video, image-to-video, and text-to-image tas...

  17. Movie Gen: A Cast of Media Foundation Models

    cs.CV 2024-10 unverdicted novelty 5.0

    A 30B-parameter transformer and related models generate high-quality videos and audio, claiming state-of-the-art results on text-to-video, video editing, personalization, and audio generation tasks.

  18. StarCoder: may the source be with you!

    cs.CL 2023-05 accept novelty 5.0

    StarCoderBase matches or beats OpenAI's code-cushman-001 on multi-language code benchmarks; the Python-fine-tuned StarCoder reaches 40% pass@1 on HumanEval while retaining other-language performance.

  19. MedThink: Enhancing Diagnostic Accuracy in Small Models via Teacher-Guided Reasoning Correction

    cs.CY 2026-04 unverdicted novelty 4.0

    MedThink, a two-stage teacher-guided reasoning correction distillation framework, boosts small language models' medical diagnostic accuracy by up to 12.7% on benchmarks and achieves 56.4% on a gastroenterology dataset.

  20. Large Language Models: A Survey

    cs.CL 2024-02 accept novelty 3.0

    The paper surveys key large language models, their training methods, datasets, evaluation benchmarks, and future research directions in the field.