HDFlow: Enhancing LLM Complex Problem-Solving with Hybrid Thinking and Dynamic Workflows

• URL: http://arxiv.org/abs/2409.17433v1
• Date: Wed, 25 Sep 2024 23:52:17 GMT
• Title: HDFlow: Enhancing LLM Complex Problem-Solving with Hybrid Thinking and Dynamic Workflows
• Authors: Wenlin Yao, Haitao Mi, Dong Yu,
• Abstract summary: We propose a novel framework HDFlow for complex reasoning with large language models (LLMs) Our approach consists of two key components: 1) a new approach for slow, deliberate reasoning called Dynamic, which automatically decomposes complex problems into more manageable sub-tasks; and 2) Hybrid Thinking, a general framework that dynamically combines fast and slow thinking based on problem complexity. Experiments on four reasoning benchmark demonstrate that our slow thinking with dynamic datasets significantly outperforms Chain-of-Thought, and hybrid thinking achieves the highest accuracy while providing an effective balance between computational efficiency and performance.
• Score: 33.035088506211096
• License: http://arxiv.org/licenses/nonexclusive-distrib/1.0/
• Abstract: Despite recent advancements in large language models (LLMs), their performance on complex reasoning problems requiring multi-step thinking and combining various skills is still limited. To address this, we propose a novel framework HDFlow for complex reasoning with LLMs that combines fast and slow thinking modes in an adaptive manner. Our approach consists of two key components: 1) a new approach for slow, deliberate reasoning called Dynamic Workflow, which automatically decomposes complex problems into more manageable sub-tasks and dynamically designs a workflow to assemble specialized LLM or symbolic reasoning tools to solve sub-tasks; 2) Hybrid Thinking, a general framework that dynamically combines fast and slow thinking based on problem complexity. Finally, we propose an easy-to-scale method for automatically synthesizing a large-scale dataset of 27K challenging reasoning problems for complex reasoning and a hybrid thinking tuning method that trains smaller LLMs on this dataset to internalize the fast/slow hybrid reasoning strategies. Experiments on four reasoning benchmark datasets demonstrate that our slow thinking with dynamic workflows significantly outperforms Chain-of-Thought, and hybrid thinking achieves the highest accuracy while providing an effective balance between computational efficiency and performance. Fine-tuning using our hybrid thinking approach also significantly boosts the complex reasoning capabilities of open-source language models. The results showcase the promise of slow thinking, dynamic workflows, and hybrid thinking in expanding the frontier of complex problem-solving with LLMs\footnote{Code and data will be released at \url{https://github.com/wenlinyao/HDFlow}.}.

Related papers

• Interactive and Expressive Code-Augmented Planning with Large Language Models [62.799579304821826]
  Large Language Models (LLMs) demonstrate strong abilities in common-sense reasoning and interactive decision-making. Recent techniques have sought to structure LLM outputs using control flow and other code-adjacent techniques to improve planning performance. We propose REPL-Plan, an LLM planning approach that is fully code-expressive and dynamic.
  arXiv  Detail & Related papers  (2024-11-21T04:23:17Z)
• Reference Trustable Decoding: A Training-Free Augmentation Paradigm for Large Language Models [79.41139393080736]
  Large language models (LLMs) have rapidly advanced and demonstrated impressive capabilities. In-Context Learning (ICL) and. Efficient Fine-Tuning (PEFT) are currently two mainstream methods for augmenting. LLMs to downstream tasks. We propose Reference Trustable Decoding (RTD), a paradigm that allows models to quickly adapt to new tasks without fine-tuning.
  arXiv  Detail & Related papers  (2024-09-30T10:48:20Z)
• DynaThink: Fast or Slow? A Dynamic Decision-Making Framework for Large Language Models [42.95876831743256]
  Large language models (LLMs) have demonstrated emergent capabilities across diverse reasoning tasks via Chains-of-Thought prompting. This paper addresses the challenge of enabling LLMs to autonomously select between fast and slow inference methods. We introduce a dynamic decision-making framework that categorizes tasks into two distinct pathways: 'Fast', designated for tasks where the LLM quickly identifies a high-confidence solution, and 'Slow', allocated for tasks that the LLM perceives as complex.
  arXiv  Detail & Related papers  (2024-07-01T06:45:13Z)
• Efficient Orchestrated AI Workflows Execution on Scale-out Spatial Architecture [17.516934379812994]
  We present "Orchestrated AIs," an approach that integrates various tasks with logic-driven decisions into dynamic, sophisticated AIs. We find that the intrinsic Dual Dynamicity of Orchestrated AIs can be effectively represented using the Orchestrated spatial Graph. Our evaluations demonstrate that significantly outperforms traditional architectures in handling the dynamic demands of Orchestrated AIs.
  arXiv  Detail & Related papers  (2024-05-21T14:09:31Z)
• Tree-of-Mixed-Thought: Combining Fast and Slow Thinking for Multi-hop Visual Reasoning [16.495754104540605]
  Large language models (LLMs) can generate code-like plans for complex inference tasks such as visual reasoning. We propose a hierarchical plan-searching algorithm that integrates the one-stop reasoning (fast) and the Tree-of-thought (slow)
  arXiv  Detail & Related papers  (2023-08-18T16:21:40Z)
• Faith and Fate: Limits of Transformers on Compositionality [109.79516190693415]
  We investigate the limits of transformer large language models across three representative compositional tasks. These tasks require breaking problems down into sub-steps and synthesizing these steps into a precise answer. Our empirical findings suggest that transformer LLMs solve compositional tasks by reducing multi-step compositional reasoning into linearized subgraph matching.
  arXiv  Detail & Related papers  (2023-05-29T23:24:14Z)
• SwiftSage: A Generative Agent with Fast and Slow Thinking for Complex Interactive Tasks [81.9962823875981]
  We introduce SwiftSage, a novel agent framework inspired by the dual-process theory of human cognition. The framework comprises two primary modules: the Swift module, representing fast and intuitive thinking, and the Sage module, emulating deliberate thought processes. In 30 tasks from the ScienceWorld benchmark, SwiftSage significantly outperforms other methods such as SayCan, ReAct, and Reflex.
  arXiv  Detail & Related papers  (2023-05-27T07:04:15Z)
• Multi-intersection Traffic Optimisation: A Benchmark Dataset and a Strong Baseline [85.9210953301628]
  Control of traffic signals is fundamental and critical to alleviate traffic congestion in urban areas. Because of the high complexity of modelling the problem, experimental settings of current works are often inconsistent. We propose a novel and strong baseline model based on deep reinforcement learning with the encoder-decoder structure.
  arXiv  Detail & Related papers  (2021-01-24T03:55:39Z)
• Optimization-driven Machine Learning for Intelligent Reflecting Surfaces Assisted Wireless Networks [82.33619654835348]
  Intelligent surface (IRS) has been employed to reshape the wireless channels by controlling individual scattering elements' phase shifts. Due to the large size of scattering elements, the passive beamforming is typically challenged by the high computational complexity. In this article, we focus on machine learning (ML) approaches for performance in IRS-assisted wireless networks.
  arXiv  Detail & Related papers  (2020-08-29T08:39:43Z)

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.