Meta Semi-Formal Reasoning: How Structured Prompting Boosts Code Review Accuracy to 93%
Meta researchers have developed semi-formal reasoning, a structured prompting technique that boosts LLM code review accuracy to 93 percent.
Tag: AI Research
AI Scientist-v2: SakanaAI’s Autonomous System Writes Peer-Reviewed Research Papers
SakanaAI releases AI Scientist-v2, an autonomous system that generates peer-reviewed research papers through agentic tree search. The AI can hypothesize, experiment, and write complete scientific manuscripts.
The AI Scientist-v2: SakanaAI Fully Autonomous Research System Wrote a Peer-Accepted Workshop Paper
SakanaAI AI Scientist-v2 generates hypotheses, runs experiments, and writes papers autonomously 鈥?and just had one accepted through peer review at an AI workshop.
Anthropic Wins First Round: Judge Blocks Pentagon’s Blacklisting of AI Company
A federal judge has granted Anthropic a preliminary injunction blocking the Pentagon's blacklisting of the AI company, ruling it may be 'classic illegal First Amendment retaliation.'
Google’s TurboQuant Algorithm Cuts AI Memory Costs by 50% ??And the Industry Is Already Using It
Google's new TurboQuant algorithm achieves 8x memory speedup and cuts AI inference costs by 50% or more. The open-source release has already been ported to llama.cpp and MLX.
Anthropic’s Claude Can Now Control Your Mac: The Future of AI Agents Is Here
Anthropic has given Claude the ability to directly control your Mac, clicking buttons and navigating apps on your behalf. This is the most ambitious step yet in the race to build AI…
last30days-skill: The AI Research Tool Taking GitHub by Storm
An AI skill that scours Reddit, X, YouTube, HN, Bluesky, and Polymarket to synthesize real-time briefings has racked up 12,000+ GitHub stars. Here's what last30days-skill does and why it's changing how developers…
AI Scientist-v2: SakanaAI’s Automated System Writes and Publishes Peer-Reviewed Research
A groundbreaking new system from SakanaAI has achieved what was once thought to be years away: an AI system that can autonomously generate scientific hypotheses, design and run experiments, analyze results, and…
DeerFlow 2.0: Bytedance’s Open-Source SuperAgent Framework Reimagines Autonomous Research and Coding
ByteDance's DeerFlow has taken the AI development community by storm, earning over 43,000 stars on GitHub and claiming the #1 spot on GitHub Trending. This revolutionary open-source framework transforms AI from a…