Unlock the principles of generative AI with insights from top Stanford experts in @StanfordOnline
’s Technical Fundamentals of Generative AI course. Master foundation models, prompt engineering, LLM optimization, and more—all with a human-centered lens: https://
stanford.io/3SEHeiA
LLMS
-
Stanford’s Technical Fundamentals of Generative AI Masterclass
By
–
-
Gemini 2.5 Pro Achieves SOTA Long Context Performance
By
–
The new Gemini 2.5 Pro is SOTA at long context, especially capable on higher number of items being retrieved (needles) as shown below!
-
How to configure custom GPTs with knowledge base files
By
–
Try it out, sure ! I would recommend to put it inside a PDF / Docx file and simply write instructions yourself for the custom GPT, referencing the knowledge base file with instructions like: “Before giving your output, strictly follow the instructions inside [NAME].docx file in
-
Sonnet 4 Model Performance Analysis and Contextual Limitations
By
–
Not nearly as good as sonnet 4 so far in my testing – it doesn’t seem to really understand the detail of how things fit together. Maybe I just haven’t found quite the right context for it.
-
AI Model Self-Optimization: Big Tech’s Efficiency Strategy
By
–
I think the theoretical scenario big tech should be leaning into is that, in the short-term only big tech can afford the most powerful models but, as the models get smarter, they should be able to solve their own efficiency issues. They'll use the intelligence to develop more
-
Smart Scaling for Efficient Language Models and Democratized AI
By
–
Senior Fellow @YejinChoinka discussed alternatives to brute-force scaling in her talk, Mission Impossible. She proposed “smart scaling” to build smaller, efficient language models that broaden access to generative AI. Watch her talk here: https://
atxsummitvp.com/mission-imposs
ible-democratising-generative-ai-by-transcending-scaling-laws/
… -
Long Context Capabilities in AI Models Beyond 192k Tokens
By
–
We like long context. Go beyond 192k plz ; )
-
Guide to using LLMs for product and service content generation
By
–
1/ Just copy paste the entire prompt. Run it with ChatGPT o3, Gemini 2.5 Pro, or Claude 4 Opus. Answer a few questions about your product / service.
-
Groq LoRA Fine-Tuning: Adapt Models Without Retraining
By
–
Your model was good. Until it wasn’t.
— Groq Inc (@GroqInc) 6 juin 2025
Performance stalled. Costs climbed.
Your team hit the limit. Again.
Groq’s LoRA fine-tuning changes that:
✅ No retraining
✅ No downtime
✅ No compromise
Now you can adapt in real time without starting over.
How it works in the comments… pic.twitter.com/1aD8tKOHhjYour model was good. Until it wasn’t.
Performance stalled. Costs climbed.
Your team hit the limit. Again. Groq’s LoRA fine-tuning changes that: No retraining No downtime No compromise Now you can adapt in real time without starting over. How it works in the comments -
ScreenSuite: Comprehensive Evaluation Suite for GUI Agents
By
–
Today we release ScreenSuite, the most comprehensive evaluation suite for GUI agents (aka Computer Use agents). We packed 13 benchmarks, and 3 different environments, to evaluate the full range of agentic capabilities for vision models. And it turns out, @Alibaba_Qwen models are