最有价值项目

A terminal UI for browsing and inspecting vector databases

ToucanDB is a brand-new micro ML-first database engine 🦜

The notebooks & projects code for Generative for Backend Dev online bootcamp

An advanced AI-powered conversational agent leveraging the Llama 3.2 model and Phidata framework. Features include reasoning, natural language interaction, and tool integration for web searches and calculations. Designed for interactivity with enhanced logging, support for custom tools, and structured outputs.

Give non-multimodal Claude Code main models the ability to see pasted screenshots — a ~200-line UserPromptSubmit hook.

CLI & async Python library for free AI chat, image & video generation.

Characterize Anything: A Wondrous Chemical Reaction between vision models and AI Characters

SemEval 2022 Task 5: Multimedia Automatic Misogyny Identification - baseline models and dataset

[EMNLP 2025 Findings] MEXA: Towards General Multimodal Reasoning with Dynamic Multi-Expert Aggregation

ACL 2024 (SRW), Official Codebase of our Paper: "MoExtend: Tuning New Experts for Modality and Task Extension"

API/MCP wrapper for RagAnything

VideoEval-Pro: Robust and Realistic Long Video Understanding Evaluation [TMLR26]

A curated list of resources for Grok, the AI developed by xAI

[NAACL 2025 Findings] Mitigating Hallucinations in Large Vision-Language Models via Summary-Guided Decoding

[ICLR 2026] Official repository of "Uni-DPO: A Unified Paradigm for Dynamic Preference Optimization of LLMs".

Recurrence Meets Transformers for Universal Multimodal Retrieval

Official Implementation of "Visual Agentic Memory: Online Long Video Understanding via Online Indexing, Hierarchical Memory, and Agentic Retrieval"

This is a multi-modal fusion method based on VGG16 and FastText for identifying useful information collected from social media platforms.

Package for bimodal training of deep neural networks on neurological data. Pypi: https://pypi.org/project/BiModNeuroCNN/

Thinking-native compiler for LEET Language Understanding item sets — killer-item design grammar, verification separation, deterministic gate chain, 3-model benchmark