Perplexity app for iOS now has R1 and o3-mini selectors as well as a new reasoning view
LLMS
-
OpenRouter launches reasoning traces with expanded inline CoT design
By
–
Reasoning traces with OpenRouter out now, with an expanded inline CoT design coming soon.
-
OpenAI Launches Advanced Research Agent and Reasoning Model
By
–
OpenAI strengthens its AI capabilities with the launch of a new research-focused agent and a more powerful reasoning model, marking significant advances in both autonomous research and model performance. Here's your daily AI news briefing for February 3rd, 2025: OpenAI
-
ChatGPT destroys DeepSeek and all AIs with Deep Research
By
–
#ChatGPT has just destroyed #DeepSeek and all AIs on the market. Really. Here is Deep Research and I am doing the full test on the channel → https://youtu.be/c-uMcNa2UXg One thing is certain: you are going to be blown away
-
Modified stronger version of surgeon riddle works on frontier models
By
–
I didn’t come up with the core idea, but I wrote a modified stronger version and showed it worked on many frontier models. Earlier examples were closer to yours — the riddle, but omitting the father’s death. Mine goes further, explicitly stating the surgeon is the father.
-
Test Jailbreaks to Improve AI System Security
By
–
Try your favorite jailbreaks on it and help us better secure powerful AI systems: https://
claude.ai/constitutional
-classifiers
… -
CodeLLM: Edit Code In-Place with AI Selection and Streaming
By
–
With our CodeLLM, you can ask AI to edit your code in-place by selecting a code block. We also have streaming responses and you can easily insert code. Try it out with ChatLLMs
-
DeepSeek Open Source AI Benefits Meta and Entire Community
By
–
It's actually very good for Meta and the rest of the open source AI world.
Since DeepSeek published the techniques and open sources the code, whatever innovation they came up with will make it to other models, including Llama.
Everyone wins.
The only parasites are the ones who -
Meta’s Open Source Strategy and AI Foundation Model Commoditization
By
–
That would be a bull case for Meta, which has 3.3 billion pairs of eyeballs on its various services. The idea that the interface to users matters and the underlying AI foundation model will become commoditized is one reason for Meta's open source strategy. And that's not
-
LLM Classification Systems Trained on Constitutional Harm Prevention
By
–
Our algorithm trains LLM classification systems to block harmful inputs and outputs based on a “constitution” of harmful and harmless categories of information.