Learning to Plan for Language Modeling from Unlabeled Data

• URL: http://arxiv.org/abs/2404.00614v2
• Date: Wed, 31 Jul 2024 12:25:14 GMT
• Title: Learning to Plan for Language Modeling from Unlabeled Data
• Authors: Nathan Cornille, Marie-Francine Moens, Florian Mai,
• Abstract summary: We train a module for planning the future writing process via a self-supervised learning objective. Given the textual context, this planning module learns to predict future abstract writing actions, which correspond to centroids in a clustered text embedding space.
• Score: 23.042650737356496
• License: http://arxiv.org/licenses/nonexclusive-distrib/1.0/
• Abstract: By training to predict the next token in an unlabeled corpus, large language models learn to perform many tasks without any labeled data. However, their next-token-prediction objective arguably limits their performance in scenarios that require planning, such as writing a coherent article. In this paper, we train a module for planning the future writing process via a self-supervised learning objective. Given the textual context, this planning module learns to predict future abstract writing actions, which correspond to centroids in a clustered text embedding space. By conditioning on these actions, our model extends the successful language model formula to more abstract planning in an unsupervised way. Empirically, we demonstrate that our method improves language modeling performance in general, particularly with respect to the text structure. Because our framework uses a planner module that is unsupervised and external to the language model, new planner modules can be trained at large scale and easily be shared with the community.

Related papers

• PARADISE: Evaluating Implicit Planning Skills of Language Models with Procedural Warnings and Tips Dataset [0.0]
  We present PARADISE, an abductive reasoning task using Q&A format on practical procedural text sourced from wikiHow. It involves warning and tip inference tasks directly associated with goals, excluding intermediary steps, with the aim of testing the ability of the models to infer implicit knowledge of the plan solely from the given goal. Our experiments, utilizing fine-tuned language models and zero-shot prompting, reveal the effectiveness of task-specific small models over large language models in most scenarios.
  arXiv  Detail & Related papers  (2024-03-05T18:01:59Z)
• CI w/o TN: Context Injection without Task Name for Procedure Planning [4.004155037293416]
  Procedure planning in instructional videos involves creating goal-directed plans based on visual start and goal observations from videos. Previous research has tackled this problem with gradually weaker training supervision, from heavy intermediate visual observations or language instructions to task class supervision. We propose a much weaker setting without task name as supervision, which is not currently solvable by existing large language models.
  arXiv  Detail & Related papers  (2024-02-23T19:34:47Z)
• PlaSma: Making Small Language Models Better Procedural Knowledge Models for (Counterfactual) Planning [77.03847056008598]
  PlaSma is a novel two-pronged approach to endow small language models with procedural knowledge and (constrained) language planning capabilities. We develop symbolic procedural knowledge distillation to enhance the commonsense knowledge in small language models and an inference-time algorithm to facilitate more structured and accurate reasoning.
  arXiv  Detail & Related papers  (2023-05-31T00:55:40Z)
• Pre-Training to Learn in Context [138.0745138788142]
  The ability of in-context learning is not fully exploited because language models are not explicitly trained to learn in context. We propose PICL (Pre-training for In-Context Learning), a framework to enhance the language models' in-context learning ability. Our experiments show that PICL is more effective and task-generalizable than a range of baselines, outperforming larger language models with nearly 4x parameters.
  arXiv  Detail & Related papers  (2023-05-16T03:38:06Z)
• Language Model Pre-Training with Sparse Latent Typing [66.75786739499604]
  We propose a new pre-training objective, Sparse Latent Typing, which enables the model to sparsely extract sentence-level keywords with diverse latent types. Experimental results show that our model is able to learn interpretable latent type categories in a self-supervised manner without using any external knowledge.
  arXiv  Detail & Related papers  (2022-10-23T00:37:08Z)
• Few-shot Subgoal Planning with Language Models [58.11102061150875]
  We show that language priors encoded in pre-trained language models allow us to infer fine-grained subgoal sequences. In contrast to recent methods which make strong assumptions about subgoal supervision, our experiments show that language models can infer detailed subgoal sequences without any fine-tuning.
  arXiv  Detail & Related papers  (2022-05-28T01:03:30Z)
• DeepStruct: Pretraining of Language Models for Structure Prediction [64.84144849119554]
  We pretrain language models on a collection of task-agnostic corpora to generate structures from text. Our structure pretraining enables zero-shot transfer of the learned knowledge that models have about the structure tasks. We show that a 10B parameter language model transfers non-trivially to most tasks and obtains state-of-the-art performance on 21 of 28 datasets.
  arXiv  Detail & Related papers  (2022-05-21T00:58:22Z)
• Text-Based Action-Model Acquisition for Planning [13.110360825201044]
  We propose a novel approach to learning action models from natural language texts by integrating Constraint Satisfaction and Natural Language Processing techniques. Specifically, we first build a novel language model to extract plan traces from texts, and then build a set of constraints to generate action models based on the extracted plan traces.
  arXiv  Detail & Related papers  (2022-02-15T02:23:31Z)
• Truth-Conditional Captioning of Time Series Data [34.65925116012727]
  We explore the task of automatically generating natural language descriptions of salient patterns in a time series. A model for this task should be able to extract high-level patterns such as presence of a peak or a dip. We propose a computational model with a truth-conditional architecture which first runs small learned programs on the input time series. We find that the proposed model is able to generate high-precision captions even though we consider a small and simple space of module types.
  arXiv  Detail & Related papers  (2021-10-05T06:28:37Z)
• Unsupervised Paraphrasing with Pretrained Language Models [85.03373221588707]
  We propose a training pipeline that enables pre-trained language models to generate high-quality paraphrases in an unsupervised setting. Our recipe consists of task-adaptation, self-supervision, and a novel decoding algorithm named Dynamic Blocking. We show with automatic and human evaluations that our approach achieves state-of-the-art performance on both the Quora Question Pair and the ParaNMT datasets.
  arXiv  Detail & Related papers  (2020-10-24T11:55:28Z)

This list is automatically generated from the titles and abstracts of the papers in this site.

This site does not guarantee the quality of this site (including all information) and is not responsible for any consequences.