the-decoder4 天
Google's DataGemma aims to ground language models in reality and curb AI hallucinations
Google has unveiled DataGemma, a set of open models designed to improve the accuracy and reliability of large-scale language models by grounding them in real-world data. According to Google, the ...
the-decoder1 个月
Microsoft releases new Phi 3.5 open-source language and vision models
Microsoft has released three new open-source AI models in its Phi series: mini-instruct, MoE-instruct, and vision-instruct. These models excel at LLM reasoning and support multiple languages, but have ...
the-decoder6 天
AI warning system Raven Sentry: How the US predicted Taliban attacks in Afghanistan
Under the pressure of dwindling resources and increasing violence in Afghanistan, the US military developed an AI system in 2019 to predict Taliban attacks. It was surprisingly accurate. As NATO ...
the-decoder1 个月
Luma AI video generator "Dream Machine" receives major update 1.5
Luma AI has released a new version of its Dream Machine video generator, which can create high-quality, realistic videos from text instructions and images. Almost two months after the first version of ...
the-decoder6 天
Code competition Codeforces bans AI code as as it reaches "new heights that cannot be ...
Faced with the increasing capabilities of AI in programming, the human-centered Codeforces coding competition has decided to ban AI entries. However, enforcing this ban may prove challenging.
the-decoder9 天
Midjourney teases Version 7, 3D system, and external image editor
During recent "Office Hours," Midjourney founder and CEO David Holz personally shared updates on current projects and answered community questions via Discord. After rapid progress in a short time, ...
the-decoder19 天
Video-01 AI video generator from MiniMax is the latest Sora challenger from China
Chinese AI startup MiniMax has unveiled Video-01, its first AI model capable of generating high-resolution videos from text prompts. CEO Yan Junjie presented Video-01 at a company event, sharing some ...
the-decoder7 天
T-FREE: Researchers develop tokenizer-free method for more efficient AI language models
With T-FREE, a research team presents a new method that does not require a tokenizer and could significantly increase the efficiency of large language models. Scientists from Aleph Alpha, the ...
the-decoder11 天
Reflection 70B launch mired in controversy as third-party benchmarks disappoint
The launch of Reflection 70B came with big promises, but third-party verification is still pending. According to the comparison platform Artificial Analysis, Reflection 70B actually underperformed in ...
the-decoder8 天
New AI model GameGen-O creates open-world video game simulations
Researchers from universities in Hong Kong and China, along with Tencent, have developed GameGen-O, an AI model that can create simulations of open-world video games. Accotding to the research team, ...
the-decoder1 个月
Nvidia researchers create AI that predicts thunderstorms with kilometer-scale precision
Nvidia researchers have developed an AI model called "StormCast" that enables thunderstorm forecasts on a scale of a few kilometers. Until now, it has been challenging for AI models to capture the ...