-
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 😉…
-
Improve your AI instruction on the fly …
Improve your AI instruction on the fly is easily doable with a simple set of generic instructions. Start as always with something basic, nothing fancy. Then move forward keeping in mind my favorite principle “First get it work, then make it better“. This will improve your instructions for any llm on its own. Below an example that works for me. Especially when i tell my LLmem that it is a nice soluion, i like it and good job. Instruction example That`s all folks ! Improve your AI instruction on the fly is nothing more then a feedback loop that is computed every time. Remember to keep overall instructions not too…
-
How to work with GitHub Copilot Limits (Without Fighting Them)
Schedule around them. GitHub Copilot is great (was better, latest updates ruined it) until you hit the quota. Then suddenly your flow breaks and it feels like someone unplugged you from the matrix. Just as you were starting to get productive and claryfing the code. Instead of treating that as a problem, I started treating it as a constraint. Nasty one, ugly one, but hey ! We did work over some worse issues. Don’t fight the limits – schedule around them. Remember to do everything You can to squeeze out as much as possible out of every token. In example : build proper context, descripe properly the plan, be precise,…
-
Greek gods in software development
Greek gods in software development can teach us. Ancient wisdom will always stand it`s ground, universally. Beyond time and space. Their domain and virtues stand firm through the ages. Relevant now as those were years ago and will be long after we are gone. Why not ? People define their own set of rules and best practices according to their own taste and work. I envourage You to do the same for Yourself. Then write those down as a proper prompts / instructions for Your AI agents so they will be good little minions. Greek gods in software development – make your own rules ! God Domain Law / Rule…
-
How to make an icon for 15 cents ?
How to make an icon for 15 cents ? Use AI. Try locally something like https://huggingface.co/models?pipeline_tag=text-to-image&sort=trending&search=ideogram This is how i made an icon for my browser extention and it cost me 15 cents. In Poland i wouldn`t buy any kind of ice cream for that and here… i have myself a nice icon for my browser extension / plugin. Best way to generate an SVG logo, for me, personally. Just use the openrouter chat option and provide some inputs, leave eveyrthing on auto. Why SVG? Cause You can easily edit it in any txt editor. It is liightweight. Scales. Cause it is text and not binary data. You can remove…
-
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…
-
Semantic collapse in RAGs by Stanford
Semantic collapse in RAGs by Stanford provides a way to keep on those MCP servers alive. After all … how do you know that you’re actually learning more, and not forgetting even more than you learn?You test ? 🙂 Will be the first one when i see an MCP server tested around the clock for proper, deterministic return values. Stanfors is the best -> dho.stanford.edu/wp-content/uploads/Legal_RAG_Hallucinations.pdf Semantics asks what an expression means. Logic asks whether an argument is valid or whether a statement follows from other statements. A sentence can be semantically meaningful but still logically false. All birds can fly. Penguins are birds. ∴ Penguins can fly. Formally valid, but…
-
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…
