✅ Please upvote my thesis please 🙏🙏
Hey guys if you just have a second can you please upvote my compute lab poll.
The in-lab version is obviously Elf SO’s but
My lab's repository focuses on Dissipation sampling of diffusion of an incursion within fractal edges derived from a cumulative centroid for the structure of the feature. The centroids magnitude extracted eigenvalue is implemented based upon a vertex mask in plural model instances to distribute reference quickly. This way we have reference in both initialization and activation. The results are applied as an n calculated embedded feature, and the curvature is applied to the outer most vertex to augment activation.
https://discord.com/channels/1325514097181327430/1325524689841291284/1378498106353647666
90 Replies
... is this a weird joke?
Bro's one step closer to making the Dimensional Merge possible and establishing Sonichu as the king of our dimension, and you're calling it a joke
My lab's repository is in easy to read C# so anyone can observe my calculations.
Find me online using: Cdascientist
Please visit my Discord Discussion and download my GitHub: Learn Ai Together
https://discord.com/channels/702624558536065165/1373816106975756431
Please upvote my poll so I can be on the show:
https://discord.com/channels/1325514097181327430/1325524689841291284/1378498106353647666
Please vote my poll 🙏🙏🙏🙏
i ain't reading all that
happy for you though
or sorry that happened
Thank you so much, if you have one second can you vote on my poll so I can talk about the lab on the podcast please 🙏: https://discord.com/channels/1325514097181327430/1325524689841291284/1378498106353647666
i can't tell if you're a bot or off your schizophrenia meds
there's... a lot of context missing here
idk if the AI stuff is just over my head or complete nonsense
this is a help forum anyway, not a beg for upvotes forum
You can look at my C# GitHub, review the controller, it’s just a compound theorem: https://github.com/cdascientist/Base_Pre
GitHub
GitHub - cdascientist/Base_Pre: Advanced ML Techniques using Advanc...
Advanced ML Techniques using Advanced Dimensional Embeddings using Spatial Coordinates and 3D Fractal Diffusion with Dimensional Coupling & Model Synthesis - cdascientist/Base_Pre
nothing you're saying makes any sense to me
and your discord links don't work, you can't link directly to channels from other servers (they only work for people already in the server)
O shit, I will put the link to my poll on my hit hub, sorry
nobody knows what the poll is, or what show you're talking about, or anything about this really
I feel ya, check out my C# TensorFlow repo
...why?
like i said, this is a help forum
if you have a C# question you can ask, this isn't a place to spam and advertise
It’s a C# lab with results, maybe read the code and digests for a second…https://github.com/cdascientist/Base_Pre
GitHub
GitHub - cdascientist/Base_Pre: Advanced ML Techniques using Advanc...
Advanced ML Techniques using Advanced Dimensional Embeddings using Spatial Coordinates and 3D Fractal Diffusion with Dimensional Coupling & Model Synthesis - cdascientist/Base_Pre
you don't seem to get it...
Ok…here is my C# question, perhaps we may defer c# scientist please:
I’m using TF and using sampling to extract eigenvectors to be placed on the outer vertices as an function. What are dangers in stability?
The rate of dissipation is the basis vector for curvature. Fractals typically being highly unstable I’m wanting to evaluate the downside, this is the same fractal every time:
float theta = (r < 1e-6f) ? 0 : MathF.Acos(z.Z / r);
float phi = MathF.Atan2(z.Y, z.X);
float newR = MathF.Pow(r, Power);
float newTheta = Power * theta;
float newPhi = Power * phi;
So we are talking about proof of concept, my objective is to circumvent the iterative nature of MLA within the compute procedure.
Deep Seek Paper, page 10 figure 3:
https://huggingface.co/papers/2412.19437
In conjunction with the strategies latest Absolute Zero:
https://huggingface.co/papers/2505.03335
What process can we use to embed low rank joint compression while at the same time providing structure for dynamic activation sequencing before MOE dispatch right? This is that correlation between establishing 0 entropy loss and providing that same context in both MOE and node activation being both context of initialization in both considerations. Essentially, is it possible to have reference in initialization, attention, and activation with some basis of reinforcement pre execution.
I am suggesting yes, with encapsulation and dynamic entropy.
In my repository I am embedding nested on the outer most vertex with a vertex mask, the basis vectors serve as both curvature and function derived from a cumulative reference before training.
This is just poc, of course operations would be on the kernel but what is cool is the ability to measure dissipation while at the same time use the rate of dissipation as a contingent to formulate curvature based upon extracted eigenvalue
yeah this is really not a place you're going to get answers to that
you need to find an AI/math discord
also, calling an 8000 line ASP.NET Core controller easy to read is hilarious
Maybe not from you or me but perhaps there are other senior .Net engineers that have had this type of work
i mean, it's not a C# question at all
The app is in C# .net
that doesn't mean your question is about C#
your question is about some AI math or something
Ok man
like, you're clearly just copying and pasting mountains of text that you already had or you're AI generating it
if you really wanted to ask a question then begging for votes on a poll is a bad start
Ok man, I just want to share the benefits of compute modeling in C#, you can download the repo and test the results yourself
why would i want to do that
if you want to share a project, go to #projects-talk and follow the guidelines there
this is #help
Its a bot bro
i think it's an actual guy
just a confused guy
Yep, and if you are a seasoned C# developer all the libraries mentioned are basic as f**
yeah it's a guy, a bot wouldn't be getting mad that i don't care about some advanced AI research advertisement
So maybe you don’t have and answer, it’s ok bro
@Cdascientist this is not an appropriate place to beg for upvotes, especially if you can't concisely summarize why we should vote for it.
i already told you i don't, i'm not in an AI research field
:kekw:
Ok, then say less
you're the one barging in and spamming a help forum with begging
if you want my advice, you'll never get anyone to look at your code taking the approach you did here
If you don’t have and answer answer then upvote and listen to the show bro
and an 8000 line controller is insanely bad structure that nobody will put in the effort to read
Ok bro, let’s see what other people say
We r just developers man, people like u we leave the science stuff to :kekw:
:PepeLaugh:
I am in agreement with @Jimmacle.
you'd think a guy as smart as this guy would know an appropriate audience for whatever this is
let’s see what other people sayan 8000 line controller is insanely bad structure that nobody will put in the effort to read
Ok, but you can use C# to also model as well, that’s why I’m sharing
the language is honestly irrelevant if your focus is the math
you need to ask AI researchers for feedback, not average C# devs
Yes, in Kernal it’s way more layered and complex, my example allows everyone to see what’s happening in Ai model activation bro
man's been cooking his brain on AI equations too long and forgets that normal people don't know anything about that
I take it you're in university or grad school? @Cdascientist publish your paper, have it peer reviewed, and email it to your professors for review. If all reviews give you a thumbs up proceed to ask the education facility you're with to publish the paper as a scientific article
Bro, it’s just a simple C# controller, all I am asking for is an upvote, then you can listen to the podcast
Its 8,000 lines
you still haven't explained why i should do any of that
Thats an entire work week to digest
do you understand that i just... don't care?
my rule of thumb of "people that use 'bro' in every sentence are impossible to argue with" continues to hold
This is like the dude at the grocery store asking me to just sign his petition for some random thing
like, you probably could have pulled me in but your behavior in this thread has been gross
@Mango , yes it’s a bit long but I could not translate from Kernal any lighter, give me a break
8000 lines is not a bit long
that's a complete lack of effort in organizing and factoring your code
Ok @Jimmacle
That’s why I put the results on the About bro
Why are you given me a hard time bro
I am giving ya a break and the benefit of the doubt. I have no clue about ya stuff I just develop things. If this is seriously something you got a passion for, I gave ya the steps people follow for this stuff. If you're not in university/grad school reach out to one saying you have a theory you want to have reviewed and published
because you're begging and spamming in my good christian C# help channel
jcotton42
my rule of thumb of "people that use 'bro' in every sentence are impossible to argue with" continues to hold
Quoted by
<@446416451230760990> from #✅ Please upvote my thesis please 🙏🙏 (click here)
React with ❌ to remove this embed.
we all know why you're here, and it's to post links to your poll and beg for votes so you can be on whatever show you're talking about but didn't even mention what show it is
:kekw:
you posted C# code in a C# help channel and think feedback about your C# code is giving you a hard time
doesn't make sense
Ok, it’s just a C# app using Accord and TF, you are messaging saying it’s too confusing. If it were an Elf file you would not be able to read the app
i don't need to understand the AI concepts to tell you that the C# code is poorly structured
Ok, then look at the results
again... why?
why should i care about your results
you haven't explained that
you're trying to advertise something, at least sell it
Because you have not taken the time to learn TensorFlow, Accord, or AG. That’s not my fault
Is it a revolutionary method of training LLMs faster? Does it make image generation hallucinate less? Does it optimize pathfinding in games?
Just give us the elevator pitch
Without using words like "transdimentsional interintegral metavalues"
Exactly, The new models use mixture of experts, to increase efficiency
I am using C sharp to demonstrate that you can send an instruction set to the mixture of experts by using a fractal
I put the entire logic of using a front door and model training together in one code base, and then separated them into steps or segments
Cool, now we know what it is in more human terms
And you need, what, upvotes? Under some poll on Discord? To get on Tonight's Show with Jimmy Fallon or something?
Yes please, the show is called This Week in Ai, and I was hoping for 30 seconds to describe the strategy
And there we have it
Should've started with that, not describing Sumerian metaphysics
i mean, shouldn't have started at all technically
because this is a help forum and not an advertising forum
@ZZZZZZZZZZZZZZZZZZZZZZZZZ fair point, I was just hoping for support from this community because I believe you should be able to use C# for ai engineering as well
next time you should probably build some rapport and present yourself a bit better to avoid getting such a negative reaction
either way, none of this is really relevant to C#, your questions are about AI/math which would be better targeted at AI/math researchers
and the C# you've given is practically unreadable, even if i did understand the AI concepts
Its ok @Jimmacle
inconsistent formatting, very long lines, no real organization
The Elf file would have been way harder to understand, I have comments, stages, sections and chapters
idk why you keep mentioning elf files, why is that relevant
as far as i'm concerned an ELF file is a linux binary
Yes, the actual laboratory implementation is all SO, Elf, and Kernal all compiled at runtime
okay, that doesn't mean your code isn't hard to read :kekw:
@Jimmacle fair
Any feedback now that you had some time to chew?
look man, i really don't know what you expect to happen here
there are no AI experts in this server that i know of, and definitely none that want to review the volume and quality of the code you've shared
you came off very poorly in this thread, so it's probably in your best interest to take a step back, make your project more presentable, and genuinely ask for opinions instead of begging to upvote your stuff
preferably in #code-review
Yeah, I didn't want to offend. I've been a dev for a better part of 2 decades, and people are dismissing C# for ai, you can be an incredible ai engineer with c#, especially with CCI, RMI, and Jit.
My hope is that combining your understanding of C# along with developing an understanding of compound theorem allows people to grow in their careers
i don't think dumping a pile of code along with a bunch of advanced AI jargon is going to help sell that view to anyone here
you came at it as an AI researcher who happened to use C#, not a C# dev making AI more accessible in the language
@Jimmacle fair point
it would be better to assume that the people here are good at C# and don't know much about the specifics of AI, not the other way around
So you want to see C# being used more for developing new models? Have you calculated the energy cost of this compared to whats currently the mainstream?
Yes, and C# is a stronger typed language. People in ai are regressing on things we do in C# already, for example the use of mathematical libraries are more direct in C#, for some reason they are using Python in a more modular state. Based upon my analysis of other ML and AI libraries they do this to try to have the ai develop solutions that already have been fully developed in C#
In Python they don't even have JIT or CCI: https://github.com/microsoft/cci
GitHub
GitHub - microsoft/cci: Common Compiler Infrastructure
Common Compiler Infrastructure. Contribute to microsoft/cci development by creating an account on GitHub.
@Jimmacle I agree completely
Concept Name: Epistemic Displacement in Asymmetric Discourse
Abstract:
Epistemic Displacement in Asymmetric Discourse refers to a communicative phenomenon in which a participant in a high-stakes or technically complex discussion introduces criticism or opposition that is epistemically misaligned with the subject matter—often fixating on superficial, aesthetic, or otherwise low-relevance aspects of the content. This displacement disrupts meaningful dialogue by substituting rhetorical contrarianism for substantive engagement.
⸻
Definition:
Epistemic Displacement in Asymmetric Discourse is the occurrence of misaligned critique within a discussion where one party engages with the subject matter at a high level of epistemic or methodological depth, while another participant, lacking similar engagement or understanding, redirects the conversation toward peripheral or stylistic concerns. This dynamic results in the erosion of discursive integrity and the marginalization of substantive content.
⸻
Diagnostic Features:
1. Asymmetry of Expertise: The interlocutors differ substantially in domain knowledge or comprehension.
2. Displacement of Focus: The dissenting party redirects attention away from the core subject to tangential or aesthetic elements (e.g., document length, tone, formatting).
3. Epistemic Non-Engagement: The critique does not interact with or acknowledge the substance of the argument or the methodology presented.
4. Performative Dissent: Opposition is expressed rhetorically, often with implied authority, despite lacking grounding in the discussed field.
5. Discursive Disruption: The original thread of inquiry or exploration is stalled or diluted due to the diversion.