headroomlabs-ai/headroom
Headroom is a context compression layer for AI agents, reducing token usage by 60-95% across tool outputs, logs, RAG chunks, files, and conversation history, while maintaining accuracy.
Awesome AI Agents › End-to-End Optimizers
xufangzhi/ENVISIONS
ENVISIONS is a cutting-edge neural-symbolic self-training framework designed to enhance the capabilities of large language models (LLMs) through an interactive evolution process. The project focuses on integrating neural and symbolic reasoning to improve LLM performance in various complex tasks, including agentic tasks, mathematical problem solving, and logical reasoning. The framework supports self-training of LLMs by leveraging environment-guided interactions, enabling models to iteratively refine their understanding and reasoning abilities. The repository provides code and resources to reproduce the self-training experiments, including scripts for training on different task domains such as MiniWob for agentic tasks, GSM for mathematical tasks, and logic tasks. The project is built to run on multi-GPU setups and utilizes advanced tools like vLLM for accelerated generation and open-instruct for LLM training. ENVISIONS is still under active development and is linked to previous work on neural-symbolism, specifically the Symbol-LLM project, which will be presented at ACL 2024. The repository includes links to the research paper, a project website, and pre-trained model checkpoints hosted on Hugging Face, covering models of various sizes (7B and 13B parameters). The framework is designed to push the boundaries of neural-symbolic integration in LLMs, aiming to improve their reasoning and problem-solving skills by combining symbolic logic with neural network learning. This approach is particularly relevant for advancing AI in domains requiring high-level cognitive functions and interactive learning capabilities. The project encourages community engagement and provides detailed instructions for setting up the environment and running training scripts, making it accessible for researchers and practitioners interested in neural-symbolic AI and LLM self-training methodologies.
https://github.com/xufangzhi/ENVISIONS
Also in End-to-End Optimizers
topoteretes/cognee
Cognee is an open-source project that provides scalable and modular memory solutions for AI agents using ECL pipelines, enabling efficient data ingestion, knowledge graph generation, and query capabilities to enhance AI performance and reduce hallucinations.
letta-ai/letta
Letta is an open-source framework for building stateful AI agents with advanced reasoning, transparent long-term memory, and context management, supporting multiple LLM backends and providing a graphical development environment.
langchain-ai/opengpts
OpenGPTs is an open-source platform that replicates and extends OpenAI's GPTs and Assistants API, offering customizable cognitive architectures, extensive tooling, and support for multiple language models and vector databases.
aiwaves-cn/agents
Agents 2.0 is an open-source framework that enables the training and self-evolution of autonomous language agents through symbolic learning inspired by neural network methodologies.
getzep/zep
Zep is an AI memory foundation that uses a temporal Knowledge Graph to enable continuous learning and personalized experiences for AI agents by efficiently managing and retrieving contextual chat and data artifacts.
algorithmicsuperintelligence/optillm
OptiLLM is an optimizing inference proxy that significantly improves the accuracy and performance of large language models on reasoning tasks without requiring any model training, by applying over 20 advanced optimization techniques during inference.
mll-lab-nu/RAGEN
A reinforcement learning framework for training multi-turn reasoning LLM agents on the StarPO algorithm, shipping ten Gym-compatible environments plus diagnostics for echo-trap and reasoning-collapse failure modes.