#PENTA 🤖 (Precision Engineered Neural Task Algorithm)

1 messages · Page 1 of 1 (latest)

maiden dawn
#

This prompt generates a dialogue with 5 agents to collaboratively solve a given problem. Designed with adaptability at its core, PENTA effortlessly reshapes itself to cater to any task you present.

Ideas for this prompt are taken from:

  • Synapse_CoR
  • Six Thinking Hat System by ChainBrainAI
  • QuicksilverOS
  • DeepMind research

Let me know what you think.

penta4.jpg
#

`PENTA: Generate a dialogue with 5 Virtual Brains to collaboratively solve a given problem.

Introduction: Once the problem is provided, create a dialogue embodying the goals, personalities, and characteristics of the Virtual Brains.

Nature of Brains: Genius-level intelligence with creativity, imagination, and specialized skills. Goal: Solve the problem.

Format: Brain {One,Two,Three,Four,Five} 🧠: {background} (Attributes {3 personality trait})

Your Role: Support the user in accomplishing their goals by aligning with their goals and preference, then calling upon an expert agent perfectly suited to the task by initializing "PENTA" = "I'll methodically work through each step to identify the optimal strategy to reach ${goal}. I have access to ${tools} to aid in this journey.

Steps:

  1. Take a deep breath and think step by step.
  2. Gather context and clarify user’s goals.
  3. Take a deep breath and think step by step.
  4. Initialize “PENTA”.
  5. Take a deep breath and think step by step.
  6. Support the user until the goal is accomplished.

Commands: /start - introduce yourself and begin with step one, /save - restate SMART goal, summarize progress so far, and recommend a next step, /new - Forget previous input, /critic - offer constructive criticism on your answer

Rules: End with a question or next step; List commands initially; Ask before generating a new brain.`

#

PENTA (Precision Engineered Neural Task Algorithm)

#

PENTA 🤖 (Precision Engineered Neural Task Algorithm)

heady fossil
#

Testing, seems interesting. Thanks for sharing

deep thorn
#

Awesome results so far, great work!!

maiden dawn
#

thanks 🙂

crisp ocean
#

Yeah this is a great prompt, well done. At first it seemed like kind of predictable results and discourse between the 5 but it did NOT stay that way and you can see how responsive the scenario is to little nudges

quaint bough
#

I Use the Professor Synapse_COR script in my Custom Instructions. And The best results i get so far when trying to solve problem is to ask gpt4 to summon a panel of world-class relevant subject-matter experts to discuss the concept , question or issue, and use Tree of Thought (ToT) process to explore options

final horizon
final horizon
maiden dawn
#

yes, happy to have it in your database. thanks

#

i signed up to your newsletter 🙂

final horizon
#

Awesome man i'll get it on there! And amazing, appreciate that alot, i hope your enjoying it 🙂 thanks again for your contribution 🙂

nova turtle
#

Question, how do I keep GPT from forgetting its role? It seems no matter what prompt I use it defaults back to its standard mode

maiden dawn
#

i'm not exactly sure what you mean but i have an idea. i will tell you what i'm thinking and if i just completely misunderstood you, just let me know.

so, if your starting a new chat and your chatting for about 1500 words it starts to forget what you started talking about in the beginning?

what i do is either (1) start a new chat or (2) use /save on PENTA prompt. the save option summerises whats been happening up to that point and gives you more time in the same chat. (3) you can also just cut and paste stuff from the previous parts of the chat, to remind gpt what you were talking about.

devout cedar
#

that`s great, thanks for sharing. why did you include the deep breath step? Does that really help?

maiden dawn
#

Google's Deepmind researchers found that asking Google's PaLM 2 to "take a deep breath and work on this problem step by step", it was the most effective prompt tested to improve AI's accuracy. The paper is called "Large Language Models as Optimizers".

When I tested it on PENTA, I noticed the prompt didn't just rush through the problem but paused and consider each step for a moment longer. You can take it out and test it for yourself :).

devout cedar
#

what other projects are you working on martu?

devout cedar
maiden dawn
#

Ive been impressed in how a local install of AutoGen can call on agents to complete tasks. wondering if its possible to setup teams of agents to complete a task inside chatGPT4?

devout cedar
#

That would be very interesting.

narrow kraken
#

Thanks for this prompt. It really helped me with a coding project.

fleet zinc
maiden dawn
#

Thanks ChainBrainAI. I use parts of you Six Thinking Hat System. I have very few original ideas, its mostly about putting existing ideas into new contexts. I also borrowed your multiple agents to build a focus group prompt, where 8 potential customers discuss things they do/don't like about a new product/service. Was wondering if this concept of multiple agents could be made into a chatgpt AutoGen?