livshitz

The answer could be in between. Who said delusion models are limited to 2d pixel generations?

if you repeat the process twice and the same exact text is written

As soon as you upload a file or add an Action, a checkbox, which is HIDEN by default, will appear, enabled. This serves as your agreement to have your conversations contribute to model training. Keep in mind, this option will only show up after you've made an upload or added an action. It was almost out of my sight, did you notice it right away?

Did you know you can make a LLM-powered app disclose its underlying instructions with a single word?

I had the opportunity to dive into the captivating realm of prompt-engineering and explore the boundaries of LLMs. I wanted to share with you the main takeaways from my article, "Exploring the Limits of Language Models: Insights from Prompt-Hacking Challenges."

What is covered:

The Wild-Llama Mini-Game: More than just fun, this mini-game is a deep dive into the world of LLMs and chatbots, offering a wealth of challenges already tackled by many. It's our way of contributing to the community.

Eye-Opening Discoveries: As you progress, the game reveals more about LLM behavior – from oversharing to fixating on certain ideas. These discoveries are quite revealing. In many cases, the shorter the malicious prompt is that more effective it is.

Security differences between GPT-4 vs. GPT-3.5: Our discussion sheds light on how GPT-4 has advanced in tackling vulnerabilities over GPT-3.5, though it's not immune to manipulation. This is also relevant for whomever creating GPTs.

Protecting LLM Applications: Highlighting the need for security, we introduce an experimental solution aimed at bolstering LLM applications against threats.

P.S. If you have any questions or thoughts about the article, feel free to share them in the comments below.

Riley Goodside's post: https://twitter.com/goodside/status/1746685366952735034

Hey HackerNews community,

Ever find yourself intrigued by the covert world of prompt injections within Large Language Models? Well, I sure did after reading Riley Goodside's enlightening post.

Fueled by curiosity, I rolled up my sleeves and crafted a nifty tool that dives deep into analyzing these texts, shedding light on the mysterious ways they operate. It's perfect for anyone looking to explore and understand the nuances of this phenomenon.

For the tech-savvy folks out there, I've included a source code reference right within the tool, so you can peek under the hood and see how it ticks.

Excited to share this creation with you all and eager to hear your thoughts!

Happy analyzing!

I've made a tool to analyze those type of texts:

https://lab.feedox.com/wild-llama/husher?input=

append the text for analysis at the end

You shouldn't adapt YAML across your stack. Instead of parsing JSON string (the output from the LLM) into an object, I'm suggesting here you parse the YAML string into an object. The article suggests that it'll be more beneficial for you to do this LLM -> YAML -> JSON in terms of time and costs.

Yes, minified JSON would be even less tokens than YAML. But: 1- LLMs tend to have very hard time to produce minified (compacted) JSON in the output, consistently. 2- As for compacted JSON input- Empirically it seems that LLMs can process it quite well for basic cognitive tasks (Information Retrieval, basic Q&A, etc), but when it comes to bit more sophisticated tasks it fails compared to exactly the same input, uncompressed. I've mentioned and provided examples in the comments of this article.

I've composed a post that could be of interest to those of you working with GPT (or any other LLM) and seeking JSON as an output. Here's a simple trick that can help reduce expenses and improve response times.

Very impressive, a star well deserved