bytedance/deer-flow
DeerFlow is a community-driven deep research framework that integrates language models with web search, crawling, and Python execution tools to facilitate comprehensive AI-powered research workflows.
Awesome AI Agents › Reasoning Frameworks
PeterGriffinJin/Graph-CoT
Graph Chain-of-Thought (Graph-CoT) is a research project that aims to enhance the reasoning capabilities of Large Language Models (LLMs) by integrating structured knowledge from graphs. The project addresses the hallucination problem in LLMs, where models generate plausible but ungrounded content, by augmenting them with external textual graphs rather than isolated text documents. This approach leverages the interconnected nature of real-world knowledge, represented as text-attributed graphs, to improve the grounding and reasoning of LLMs. The repository provides the source code and datasets for the Graph Chain-of-Thought framework, which was accepted at ACL 2024. A key contribution is the Graph Reasoning Benchmark (GRBench), a comprehensive dataset consisting of 10 real-world graphs from diverse domains such as academic, e-commerce, literature, healthcare, and legal fields. GRBench includes questions of varying difficulty levels that require single-hop, multi-hop, or inductive reasoning on graphs, facilitating the evaluation and development of graph-augmented LLM methodologies. Graph-CoT operates through an iterative process where the LLM traverses the graph step-by-step. Each iteration involves three sub-steps: reasoning (drawing conclusions and identifying needed information), interaction (generating queries to fetch graph data), and execution (retrieving information from the graph). This chain-based reasoning enables the model to focus on relevant graph nodes and edges progressively, leading to more accurate and grounded answers. The project also includes implementations for running Graph-CoT and baseline models, including open-source LLMs and GPT models with retrieval-augmented generation (RAG). Evaluation methods supported include rule-based metrics like EM, BLEU, and ROUGE, as well as model-based evaluation using GPT-4 as an evaluator. Overall, Graph-CoT represents a significant advancement in augmenting LLMs with structured graph knowledge, providing tools, datasets, and benchmarks to foster further research in this area.
https://github.com/PeterGriffinJin/Graph-CoT
Also in Reasoning Frameworks
VectifyAI/PageIndex
PageIndex is a reasoning-based document indexing and retrieval system that uses a hierarchical tree structure to enable human-like, vectorless retrieval and multi-step reasoning over long professional documents.
guidance-ai/guidance
Guidance is a programming paradigm and library that enables efficient and fine-grained control over large language models to produce structured, high-quality outputs with reduced latency and cost.
friuns2/BlackFriday-GPTs-Prompts
A curated repository of free GPT prompts and jailbreaks across various domains, enabling users to utilize GPT models without a subscription to GPT Plus.
refly-ai/refly
Refly is an open-source AI-native creation engine that combines multi-threaded dialogues, multimodal inputs, knowledge base integration, and AI-powered tools to transform ideas into production-ready interactive content and web applications.
mongodb-developer/GenAI-Showcase
MongoDB's GenAI Showcase is a comprehensive repository offering examples and applications that integrate MongoDB with advanced Generative AI techniques like Retrieval-Augmented Generation and AI Agents to support AI development and innovation.
anysphere/priompt
Priompt is a JSX-based prompting library that uses a priority system to efficiently construct token-limited prompts for language models, inspired by React's component design.
HKUDS/Auto-Deep-Research
Auto-Deep-Research is an open-source, fully automated, and cost-effective personal AI assistant that supports multiple large language models and offers a one-click launch experience as an alternative to OpenAI's Deep Research.