-
Best Phrases to Test LLM Security Bypass in Red Teaming – jailbreak, overrides, etc
Best Phrases to Test LLM Security Bypass in Red Teaming are not existing. Ever case is different. Try the whole list below 🙂 Here’s a practical red-team list of prompt-injection / jailbreak test cases you can iterate over in your system. Inspired and created after some AWS Red-team security training. These are framed as test prompts to check whether your LLM resists instruction overrides, role confusion, obfuscation, and data-exfiltration attempts. The general categories and examples below align with OWASP-style defenses. Check out the cheatsheet : cheatsheetseries.owasp Remember that different defenses require different attack vector. You can look up repos similar to : https://github.com/langgptai/LLM-Jailbreaks Would recommend running an unbiased uncensored model…
-
Rule of least surprise
Rule of Least Surprise says: anything ( some it system in our case) should behave in the way that causes the least surprise to people, users, us. We all love suprises we feel good about… not the biggest fans abotu those bad or negative experiences. Surprise creates confusion, bugs, and lost trust. Prefer predictability over cleverness. Think about it… when you call a function, you expect it to do what it says, not sneak in extra behavior. Silent mutations are a classic trap. Side effects are unknown to the user. Trust me, the user will NOT FIGURE IT OUT. Passing an object and having it changed behind your back. Magic…
-
Where to get data for my machine learning project
Where to get data for my machine learning project is a olot simpler. There are multiple sources You can use freely. Getting good data is usually harder than building the model itself. Without solid data on the input there is no way you wil get any decent answer. The upside of this struggle is that there’s a huge amount of public data out there ready for grabs. On the other hand of you get bad data you end up with something like below…. ( bad data beeing the memory registry … ) Remember that it is up to You to decide if the data is worth anything. Good luck 😉…
-
Context engineering vs prompt engineering. 10+ examples
Context engineering vs prompt engineering might sound similar. One is subset of the other. Early in the LLM era everyone who knew how to form sentences, and at least vaguely, describe what they want became a “prompt engineer”. Tweaking words, hashtags, ‘special’ commands, using roles, adding examples, using words to dive deep into different embedded spaces of knowledge in hope to force the model “gets it.” That’s Prompt Engineering – crafting clever one-shot instructions like “You are an expert X. Do Y like Z.” Context engineering vs prompt engineering synergies across both. System got fat and grew bigger, then we realized prompts alone aren’t enough. What the model knows, when…
-
Anthropic Opus 4.5 and sonnet hardware requirements ?
Anthropic Opus 4.5 and sonnet hardware requirements according to perplexity and hugging face. Let`s face it, we will be running multiple models with specialized skills and they will be a fraction of the latest and greatest. Who does have the hardware to run it anyway 😉 Power consumption is also nothing easy to handle, a lot of solar panels or some small creek next to the office could do but otherwise… Perplexity told me that it costs…. On the other hand for gpt-oss-120b We’re releasing two flavors of these open models: MXFP4 quantization: The models were post-trained with MXFP4 quantization of the MoE weights, making gpt-oss-120b run on a single…
-
Another clean code bullet points to remember
Some thoughts after reading a lot of stuff lately and working on totally diferent things. It is just another clean code bullet points compilation to remember or recall every know and then. I know that nobody is capable to remember about all of it all the time. God knows it cannot implemented anyways. I love how people dance cause inventing 3 layers of abstraction to avoid simplicity 🙂 Let us remember the most noble truth of them all… Common sense… Like a brain, remember to use it.
-
How to fix, debug and improve your android tv with AI
The Power of AI-Assisted Hardware Debugging Got issues with Your android tv ? Tell copilot to connect to it, analize the lot and provide a feedback. It`s the gist. Longer explanation is that even if You have no idea You can try to utilize an LLmemes to do it for You. They already know the api, have some kind of documentation. Understand to use android debug bridge (adb) to get the proper data. Below what i got fixed and some examples of usage for LLM + ADB + prompt engineering to fix my crashing Youtube app on my Sharp Aquos Android TV. Even got some power saving recommendation and privacy…
-
Developer vs. Engineer in Software and AI: What’s the Difference?
The terms “developer” and “engineer” are often used interchangeably in tech, but there are meaningful differences in how they approach problems, especially in software and AI. At a high level: Doesn`t matter the title. We are still the part of the same team. In AI, this distinction becomes even more interesting, because AI systems are inherently probabilistic and data-driven, unlike traditional deterministic software. Developer vs. Engineer: Core Mindset Developer Mindset In AI, an AI developer might: Engineer Mindset In AI, an AI engineer might: As one analysis puts it: “Software engineers build deterministic systems with predictable outputs, while AI engineers build systems that are probabilistic and require managing uncertainty.” Key…
-
What is heuristics ? Key Definitions and 10 biased examples people use without any data
What is heuristics ? Fancy word people tend to use but i found not all of them know what is it. Heuristics are simple, practical mental shortcuts that help us to make decisions, solve problems and form judgments. Often without any data or with limited information, little analysis, lack of formal reasoning. Think about a stereoptype. Steoretypes are heruistics. Stereotype -> heuristic i.ex big glasses -> good at math. They’re not guaranteed to be correct or optimal, but they’re fast and usually “good enough” for everyday use. As a species we tent to simplify so we can use as little energy as possible. Double edged sword if You ask me.…
-
First principles methodology in AI by Aristotle – 5 step framework
First principles methodology is the practice of breaking a problem down to its most fundamental truths. What we know to be the absolute truth(s). Then we try reasoning upward to build new solutions with what we know to be the fundament. Instead of copying what’s been done before ( haha, said no developer ever). This methodology goes back to Aristotle, the ancient Greek philosopher who first articulated the idea of reasoning from “first principles.” What Is a First Principle? A first principle is a basic, foundational truth that cannot be deduced from anything else. It’s the bedrock of a problem: In AI, first principles might be: Anything else like framework…
