Tag
I built an open-source instrument for exploring LLM internals combining research on jspace, j++, LARQL/VINDEX, and Cammarata's neuron poles, released as logos-theodolite on GitHub with j-lenses originally fitted for gemma4-31b.
HIVE (Hive Inspired Virtual Entity) 是一个开源研究系统，通过九个从不同角度分析的 forager agent、一个 queen verdict 以及共享记忆（comb），并利用 Unpaywall 与 DOI 检索的已验证白皮书，为 LLM 输出提供可信度置信值，减少幻觉。系统可在本地运行（Ollama + Qwen3.6 35B），在 M1 Max 64GB 上设计与测试。
DEDA is an open-source Python toolkit for extracting, decoding, and anonymizing printer tracking dots (yellow dots) embedded in colour laser printouts, developed from TU Dresden research published at ACM IH&MMSec '18.
Where Is the Planet is a web tool that predicts the current locations of directly imaged exoplanets and brown dwarf companions for use in observational astronomy research, built by Jason Wang, Matas Kulikauskas, and Sarah Blunt at Northwestern.
OpenScience, a scientific agent, has launched out of beta on Product Hunt with a redesigned IDE, integrated AI models, autoresearch capabilities, and is free and open source for research labs.
Baidu's Kooko AI reached over 10 million international users in its first year, offering an integrated AI workspace for streamlined research and professional report generation.
Jev Reviewer is an open-source tool that assists systematic reviewers by using AI to extract direct quotes from research articles, supporting various file types and ensuring all processing occurs locally in the browser.
The tweet praises the camera feature in Google Notebook mobile app for enhancing on-the-go research and suggests adding bots to improve it, urging more resource allocation.
The author has built Global Fail Map, an open-source interactive tool documenting historical failures in science, technology, and business to serve as a research layer for AI agents and developers.
Atria Dawn Preview is a research agent from Shanghai that verifies its own work using external evidence to ensure verifiable and reproducible outcomes for long-horizon tasks.
ZipBench presents a low-cost framework for compressing large language model benchmarks, reducing evaluation expenses while maintaining accuracy, and introduces ZipBench Zoo with compact benchmark versions.
AgentsDock is a multi-platform IDE for agentic AI research that enables agent management, server connectivity, and integrated terminal access.
Google DeepMind's AlphaGenome Atlas is an AI-powered map of all possible single-letter DNA mutations, providing a 1-petabyte dataset for researchers to explore genetic variants via a free web interface and API.
AlphaGenome Atlas is an AI model that predicts the impact of all 9 billion possible single-letter DNA variants to aid in understanding human biology and disease.
The author introduces ARC-AGI-N, an open-source AI research tool that helps users find unsolved math and science problems and prepare prompts for agents to solve them.
The user utilized Story Prism's new Agentic-Powered Build Mode to efficiently connect 178 sources from the Epstein files, uncovering a disturbing pattern in Epstein's intellectual network and its potential implications for societal control.
Grok Bot transforms X into a conversational research database by allowing users to read and analyze timelines, mentions, likes, and trends without posting capabilities.
Deep Research is an AI agent that generates citation-rich, analyst-grade reports from open-ended questions, featuring source evaluation and multi-round investigation across hundreds of sources.
A research tool website designed for long-term investors. It tracks stock-related documents and extracts key insights while building counterarguments to deepen analysis.
Lophius is a new hybrid code/GUI research system for language models that runs inside a notebook, aiming to reduce boilerplate and streamline tasks like model inspection, tokenizer analysis, and inference.