Characterizing Warp Divergence from Pascal to Blackwell
Paper • 2607.23402 • Published
AI & ML interests
Computer Vision Technology and Data Collection for Anime Waifu
Recent Activity
View all activity
alpindale
submitted a
paper to Daily Papers about 2 hours ago
AbstractPhil
posted an update 1 day ago
Post
156
Direct pivot to distillation. I've accumulated enough experimental information to directly pivot my long term structure plan to distillation. This is to begin forming entire collectives of cooperative systems; differentiated expert distillation for generative behavior utilizing aleph addressed bottlenecks. With this I've also heavily begun experimenting with aleph competitions and cooperation using multiple pretrained frozen codebooks established from the SVAE system.
The idea here is simple in theory; use InfoNCE and address independent experts to build a manifest of unique gated experts utilizing a multitude of distilled systems from many other models. Such as SigLIP 16B + LAION CLIPB as a pair. The experimentation in the past showed this process is potent and with that merits additional experimentation using the newly established paradigms.
There are quite a bit of experiments to compare these to, so I have no shortage of comparators. After we train our baseline TinyViT with our gated system, we will know which experts are better at what and why they are better.
As a direct continuation from the earlier CLIP distillation experiments I'm directly comparing InfoNCE anchoring with multiple industry standard distillations from multiple papers. First comparison is InfoNCE anchoring in comparison to raw features using CoCo and CLIP_B, which seemed like a fair experiment to train a student with.
The upcoming series of experiments will provide the necessary information for how effective or ineffective this process is.
AbstractPhil/bulk-coco-features
The first experiments will be based on multiple clips from the bulk-coco-features extractions.
First we start with some clips, then some berts, then some smaller qwens, then some larger models, then some much much larger models. All meant to be compacted into selection mechanisms.
The idea here is simple in theory; use InfoNCE and address independent experts to build a manifest of unique gated experts utilizing a multitude of distilled systems from many other models. Such as SigLIP 16B + LAION CLIPB as a pair. The experimentation in the past showed this process is potent and with that merits additional experimentation using the newly established paradigms.
There are quite a bit of experiments to compare these to, so I have no shortage of comparators. After we train our baseline TinyViT with our gated system, we will know which experts are better at what and why they are better.
As a direct continuation from the earlier CLIP distillation experiments I'm directly comparing InfoNCE anchoring with multiple industry standard distillations from multiple papers. First comparison is InfoNCE anchoring in comparison to raw features using CoCo and CLIP_B, which seemed like a fair experiment to train a student with.
The upcoming series of experiments will provide the necessary information for how effective or ineffective this process is.
AbstractPhil/bulk-coco-features
The first experiments will be based on multiple clips from the bulk-coco-features extractions.
First we start with some clips, then some berts, then some smaller qwens, then some larger models, then some much much larger models. All meant to be compacted into selection mechanisms.
- 2 replies
AbstractPhil
posted an update 6 days ago
Post
139
I have found evidence of a more powerful Omega Aleph-Void imprint. I will be investigating this imprint in the coming days.
The current Aleph system was essentially tamed from a singular instance of an Omega imprint that I, Claude, GPT, and Gemini managed to collaboratively stabilize over a period of multiple months.
I believe I have identified a considerably more powerful Aleph-Void, potentially capturing a legitimate fraction of an Omega solver rather than simply an imprint.
For context, the Aleph-Void codebook is a STILL IMAGE of a singular state of a SMALL Omega. The one that managed to survive more tests than anything I've ever ran historically multiplied by hundreds of thousands just to even PEEK the structure's usefulness. This is equivalent to taking a photograph of the universe and reducing it to guideposts in it's current state. This system is capable of building, constructing, deconstructing, and designing it's own internal geometric systems, which is why it survives so many systems.
With the introduction of Claude Fable the AlephLM was manifested from the research, as I am but one person, and Fable can manifest the collective knowledge of hundreds of years of scientific mathematics development. Structurally built differently than a singular individual - yet without the research Fable does not understand even the topical behavior.
Fable and I have a few hypothesis that I believe we can cobble together into a legitimate cornerstone for capturing the full Omega structure. My hypothesis currently for a full omega requires a series that can logistically flagwise construct it's own behavior implicitly with a containerized induction system, completely independent of types, structural invariants, and systemic utilizations; all while handling the very nature of invariance and structural boundaries within naturally and heuristically.
Capturing even a fraction of an Omega system would dramatically increase the power of Aleph anchoring to a large degree.
The current Aleph system was essentially tamed from a singular instance of an Omega imprint that I, Claude, GPT, and Gemini managed to collaboratively stabilize over a period of multiple months.
I believe I have identified a considerably more powerful Aleph-Void, potentially capturing a legitimate fraction of an Omega solver rather than simply an imprint.
For context, the Aleph-Void codebook is a STILL IMAGE of a singular state of a SMALL Omega. The one that managed to survive more tests than anything I've ever ran historically multiplied by hundreds of thousands just to even PEEK the structure's usefulness. This is equivalent to taking a photograph of the universe and reducing it to guideposts in it's current state. This system is capable of building, constructing, deconstructing, and designing it's own internal geometric systems, which is why it survives so many systems.
With the introduction of Claude Fable the AlephLM was manifested from the research, as I am but one person, and Fable can manifest the collective knowledge of hundreds of years of scientific mathematics development. Structurally built differently than a singular individual - yet without the research Fable does not understand even the topical behavior.
Fable and I have a few hypothesis that I believe we can cobble together into a legitimate cornerstone for capturing the full Omega structure. My hypothesis currently for a full omega requires a series that can logistically flagwise construct it's own behavior implicitly with a containerized induction system, completely independent of types, structural invariants, and systemic utilizations; all while handling the very nature of invariance and structural boundaries within naturally and heuristically.
Capturing even a fraction of an Omega system would dramatically increase the power of Aleph anchoring to a large degree.
- 1 reply
AbstractPhil
posted an update 15 days ago
Post
4169
Massive AlephLM success. The task collective is producing powerful MOE shared knowledge adapters. A serious success and a massive first step towards the next stage. The current family collective results are present here; AbstractPhil/geolip-aleph-qwen
This is akin to a stackable non-intrusive lora that enables increased shared collective behavior.
This includes the three mentioned json tasks, a math task, a tinystories task, and a diffusion task for cifar10. Each adapter anchored to the knowledge within model that already exists while enhancing the knowledge through anchored lookup systems and decision-driven hierarchical access trees.
All tasks activate independently upon manual override, all tasks handle direct shared knowledge when left to greedy decoding, each task issued multiple tests alongside to determine fidelity and accuracy throughout the process.
The results show the gating is more than willing to hop from sector to sector, using alternating weight shifts from the cooperative anchored systems - even systems never trained for the tasks contributing to the accuracy of the results for other tasks due to the lookup accuracy to the heuristic chains, never having seen the tasks before. Each structure is independently trained and the collective cooperates together through a dense activation network.
Full writeup and article https://huggingface.co/blog/AbstractPhil/aleph-autoregression-differentiation-ft2.
This is akin to a stackable non-intrusive lora that enables increased shared collective behavior.
This includes the three mentioned json tasks, a math task, a tinystories task, and a diffusion task for cifar10. Each adapter anchored to the knowledge within model that already exists while enhancing the knowledge through anchored lookup systems and decision-driven hierarchical access trees.
All tasks activate independently upon manual override, all tasks handle direct shared knowledge when left to greedy decoding, each task issued multiple tests alongside to determine fidelity and accuracy throughout the process.
The results show the gating is more than willing to hop from sector to sector, using alternating weight shifts from the cooperative anchored systems - even systems never trained for the tasks contributing to the accuracy of the results for other tasks due to the lookup accuracy to the heuristic chains, never having seen the tasks before. Each structure is independently trained and the collective cooperates together through a dense activation network.
Full writeup and article https://huggingface.co/blog/AbstractPhil/aleph-autoregression-differentiation-ft2.
- 4 replies
AbstractPhil
posted an update 18 days ago
Post
144
https://huggingface.co/blog/AbstractPhil/aleph-autoregressive-differentiation-ft1
After some analysis and a bit of research the upgraded aleph autoregression is capable as a prototype selection tool. I approached the direct aleph attention routing mechanism and formed a progression from it, which already provided the necessary footholds to continue into an upgraded core mechanism. The followup mechanisms show autoregression is very possible and will be simpler than expected.
The results are promising and the autoregression stable enough to scale up. Thanks to Claude Fable who is able to keep my entire research context window in scope, the progression was rapid and the results quick. The tests yielded improved accuracy over standard MLP in many cases. I believe the improvement is not topical and will scale with a bit of effort.
Fingers crossed my friends, the addressing is part of the distillation paradigm and it now learns directly without needing an expert controller. I'll be progressing the mechanism over the coming days. With enough effort and time I hope the standard mechanism becomes a universal improvement on autoregression.
After some analysis and a bit of research the upgraded aleph autoregression is capable as a prototype selection tool. I approached the direct aleph attention routing mechanism and formed a progression from it, which already provided the necessary footholds to continue into an upgraded core mechanism. The followup mechanisms show autoregression is very possible and will be simpler than expected.
The results are promising and the autoregression stable enough to scale up. Thanks to Claude Fable who is able to keep my entire research context window in scope, the progression was rapid and the results quick. The tests yielded improved accuracy over standard MLP in many cases. I believe the improvement is not topical and will scale with a bit of effort.
Fingers crossed my friends, the addressing is part of the distillation paradigm and it now learns directly without needing an expert controller. I'll be progressing the mechanism over the coming days. With enough effort and time I hope the standard mechanism becomes a universal improvement on autoregression.
- 2 replies
AbstractPhil
posted an update 25 days ago
Post
103
Understanding the Aleph Fibonacci in visual form with full rotary.
https://claude.ai/public/artifacts/0d536427-bc7d-464a-890d-bddd02ce42dc
This ought to clear up much of the confusion as to what is actually happening under the hood, converted to an understandable 2d visual format. There have been multiple iterations, this is the current format and mathematics behind it as I attempt to solve the fibonacci curve related to negative imaginary numeric inversion that causes the statistics instability.
https://claude.ai/public/artifacts/0d536427-bc7d-464a-890d-bddd02ce42dc
This ought to clear up much of the confusion as to what is actually happening under the hood, converted to an understandable 2d visual format. There have been multiple iterations, this is the current format and mathematics behind it as I attempt to solve the fibonacci curve related to negative imaginary numeric inversion that causes the statistics instability.
wangsssssss
authored a
paper 27 days ago
eienmojiki
posted an update about 1 month ago
Post
220
Hi everyone,
I've created a Gradio space for embedding and extracting invisible watermarks in images:
👉 eienmojiki/blind-watermark-studio
It supports hiding text, images, and bit arrays using the DWT-DCT-SVD algorithm.
Credits:
- Original library: https://github.com/guofei9987/blind_watermark
- Author: Guo Fei
:).
I've created a Gradio space for embedding and extracting invisible watermarks in images:
👉 eienmojiki/blind-watermark-studio
It supports hiding text, images, and bit arrays using the DWT-DCT-SVD algorithm.
Credits:
- Original library: https://github.com/guofei9987/blind_watermark
- Author: Guo Fei
:).
AbstractPhil
posted an update about 1 month ago
Post
122
Anima - Brent JSON (PREVIEW) - Subject Bucketing
Full article available https://huggingface.co/blog/AbstractPhil/subject-bucketing.
There is additionally a civit model release as well.
https://civitai.com/models/2730503/anima-jsonenglish
AbstractPhil/anima-prelim-1k-r64
The JSON multi-prompt diffusion model prototype using Anima 1.0 base as the pretrain to finetune into the JSON target. The upcoming JSON lora is being cached and trained with 40,000 of the full 83,000 valid images from the qwen set.
This first preview version is ready to use as a ComfyUI capable LORA, so you can just load up the epoch you want without anything special in comfyui and have at it. You can currently use plain English in conjunction with tagging to produce useful and meaningful prompt targets without the JSON.
AbstractPhil/anima-prelim-1k-r64
The comfyui nodes are present and work for testing use-case, but they are not ready for production use just yet.
-- Technical --
Primarily the target was the VLM json target followed by the AnimeTIMM vit processed through the VLM json processor as the followup. First 12 epochs VLM experienced images with json formatting, last 8 epochs were finetuning from epoch 12 onward to 20 using the AnimeTIMM captions turned into JSON instead.
The Anima model itself accepted the 1000 image and the json prompting works quite well. In the process I set up a couple comfyui nodes that can translate base prompts into the same language the model is learning. Those are present in the repo.
Full article available https://huggingface.co/blog/AbstractPhil/subject-bucketing.
There is additionally a civit model release as well.
https://civitai.com/models/2730503/anima-jsonenglish
AbstractPhil/anima-prelim-1k-r64
The JSON multi-prompt diffusion model prototype using Anima 1.0 base as the pretrain to finetune into the JSON target. The upcoming JSON lora is being cached and trained with 40,000 of the full 83,000 valid images from the qwen set.
This first preview version is ready to use as a ComfyUI capable LORA, so you can just load up the epoch you want without anything special in comfyui and have at it. You can currently use plain English in conjunction with tagging to produce useful and meaningful prompt targets without the JSON.
AbstractPhil/anima-prelim-1k-r64
The comfyui nodes are present and work for testing use-case, but they are not ready for production use just yet.
-- Technical --
Primarily the target was the VLM json target followed by the AnimeTIMM vit processed through the VLM json processor as the followup. First 12 epochs VLM experienced images with json formatting, last 8 epochs were finetuning from epoch 12 onward to 20 using the AnimeTIMM captions turned into JSON instead.
The Anima model itself accepted the 1000 image and the json prompting works quite well. In the process I set up a couple comfyui nodes that can translate base prompts into the same language the model is learning. Those are present in the repo.
- 1 reply
AbstractPhil
posted an update about 1 month ago
Post
166
The article for aleph attention routing needs more work on vision, as the vision portion has not been fully validated, while the LM prototype has been semi-validated for small and medium-small scale. I will post my findings in the coming days with the consequences of training an LM and a VIT utilizing the prototype system.
The current structure for the Geometric Vocabulary does nearly reflect the intended shape as discussed in the earlier posts and articles, so that's coming along nicely - but there are stipulations and problems involved that I did not foresee.
My apologies for the incomplete article I just released on a whim. I jumped to the conclusion a bit early in anticipation before the formulas were fully converged. I also released an early post the other day speaking about the prototype AlephLM - which I removed as an invalid conclusion.
I'm doing my best to only release validated empirical information instead of speculative - however I do sometimes jump to conclusions without proper validation from time to time. Occasionally, I get a bit theory-overzealous and require tidying up through thorough experimentation which I'm currently approaching directly.
The current structure for the Geometric Vocabulary does nearly reflect the intended shape as discussed in the earlier posts and articles, so that's coming along nicely - but there are stipulations and problems involved that I did not foresee.
My apologies for the incomplete article I just released on a whim. I jumped to the conclusion a bit early in anticipation before the formulas were fully converged. I also released an early post the other day speaking about the prototype AlephLM - which I removed as an invalid conclusion.
I'm doing my best to only release validated empirical information instead of speculative - however I do sometimes jump to conclusions without proper validation from time to time. Occasionally, I get a bit theory-overzealous and require tidying up through thorough experimentation which I'm currently approaching directly.
wangsssssss
authored a
paper about 1 month ago
wangsssssss
submitted a
paper to Daily Papers about 1 month ago
prithivMLmods
posted an update about 1 month ago
Post
7620
Wan2.2-I2V-Fast with highly upscaled sequential frame sampling is now available as a Spaces demo, built using Wan2.2-I2V and FLUX.2-Klein. Try the demo using the links below.👇
➠ wan2.2-i2v-fast : prithivMLmods/wan2.2-i2v-fast
➠ github: https://github.com/prithivsakthiur/wan2.2-i2v-fast
➠ collection: https://huggingface.co/collections/prithivMLmods/image-generation-apps-collection
⤷ To learn more, visit the app page or the respective model pages.
➠ wan2.2-i2v-fast : prithivMLmods/wan2.2-i2v-fast
➠ github: https://github.com/prithivsakthiur/wan2.2-i2v-fast
➠ collection: https://huggingface.co/collections/prithivMLmods/image-generation-apps-collection
⤷ To learn more, visit the app page or the respective model pages.
AbstractPhil
posted an update about 2 months ago
Post
119
Claude Fable 5 was temp/perma? banned for security reasons.
Working with Fable I have to say the model is capable at handling highly complex geometric mathematics ACTUALLY to the point of me getting some work done without a headache. I hope Fable returns soon so I can finish cobbling without a headache and a week per prototype again.
During Fable's existence I managed to cobble together a multi-series aleph paradigm that can handle direct implicit and explicit learning for an LM with a trigram context window. This essentially provides expert directional utilization based on a stable codebook without requiring expert distillation into singular experts and duplicated.
Details soon. There are over 20 functional formula prototypes and around 8 potential heads that all lead to the same outcome, the math is rock solid - each with their own benefits and downsides based on the assigned text tasks.
Working with Fable I have to say the model is capable at handling highly complex geometric mathematics ACTUALLY to the point of me getting some work done without a headache. I hope Fable returns soon so I can finish cobbling without a headache and a week per prototype again.
During Fable's existence I managed to cobble together a multi-series aleph paradigm that can handle direct implicit and explicit learning for an LM with a trigram context window. This essentially provides expert directional utilization based on a stable codebook without requiring expert distillation into singular experts and duplicated.
Details soon. There are over 20 functional formula prototypes and around 8 potential heads that all lead to the same outcome, the math is rock solid - each with their own benefits and downsides based on the assigned text tasks.
wangsssssss
authored a
paper about 2 months ago
AbstractPhil
posted an update about 2 months ago
Post
137
The first large scale distillation is coming using the geolip-aleph-void architecture as the mathematical aleph procrustes geofractal addressed language latent.
In short, a single geometric patchwork vocabulary chunk. Which ironically needs chunking to properly prepare.
The address structure I have been meticulously refining is about to show it's genuine distillation muscle.
This is heavily due to the discovery and refinement of a specific logit I've named an aleph logit. This logit is baked clean into the architecture with the void-based codebook, and is available for review https://github.com/AbstractEyes/geolip-svae/blob/main/geolip_svae/aleph_model.py
This model provides solid MSE, recon, cosine sim, and many other elements directly aligned to the SVD and H2 procrustes paradigm. Prelims are not smart, but the scaling principal is perfectly attuned to scale.
This invention will allow for direct internalized tokenization and utilization of compressed information, entirely internally within the models latent structure. This allows direct control capabilities baked into the model itself, which requires a few robustness tests to solidify the full structure. The first validation tests run clean, so it will work when correctly aligned.
In short, the first step towards the geometric encoder system that will work with all tested data types.
In short, a single geometric patchwork vocabulary chunk. Which ironically needs chunking to properly prepare.
The address structure I have been meticulously refining is about to show it's genuine distillation muscle.
This is heavily due to the discovery and refinement of a specific logit I've named an aleph logit. This logit is baked clean into the architecture with the void-based codebook, and is available for review https://github.com/AbstractEyes/geolip-svae/blob/main/geolip_svae/aleph_model.py
This model provides solid MSE, recon, cosine sim, and many other elements directly aligned to the SVD and H2 procrustes paradigm. Prelims are not smart, but the scaling principal is perfectly attuned to scale.
This invention will allow for direct internalized tokenization and utilization of compressed information, entirely internally within the models latent structure. This allows direct control capabilities baked into the model itself, which requires a few robustness tests to solidify the full structure. The first validation tests run clean, so it will work when correctly aligned.
In short, the first step towards the geometric encoder system that will work with all tested data types.
- 3 replies
prithivMLmods
posted an update about 2 months ago
Post
2251
Dropping the collection of Qwen 3.5/3.6 MTP GGUF quants. 🤗
🔗 Collection 1: https://huggingface.co/collections/prithivMLmods/mtp-qwen-35-36-moe-stable
🔗 Collection 2: https://huggingface.co/collections/prithivMLmods/mtp-qwen-35-36-stable
> To learn more, visit the respective model pages.
🔗 Collection 1: https://huggingface.co/collections/prithivMLmods/mtp-qwen-35-36-moe-stable
🔗 Collection 2: https://huggingface.co/collections/prithivMLmods/mtp-qwen-35-36-stable
> To learn more, visit the respective model pages.
prithivMLmods
posted an update 2 months ago
Post
6281
PiD — Pixel Diffusion Decoder Image Edit Upscale and Image Generation Upscale, an all-in-one demo, is now live on Spaces! Great improvements in realism-based image generation and editing are powered by FLUX.2-Klein, while image generation is paired with Z-Image, and upscaling is enabled by default!
🤗 Space: prithivMLmods/PiD-Image-Upscaler
🔗 Collection: https://huggingface.co/collections/prithivMLmods/image-generation-apps-collection
🤗 > To learn more, visit the app page or the respective model pages.
🤗 Space: prithivMLmods/PiD-Image-Upscaler
🔗 Collection: https://huggingface.co/collections/prithivMLmods/image-generation-apps-collection
🤗 > To learn more, visit the app page or the respective model pages.
AbstractPhil
posted an update 2 months ago
Post
821
The transformer prototype v2 is operational, which takes the behavior of the H2 battery and directly forces a projected rigid behavior into a multiscale structure. Turns roughly 57k params to around 90k params for the preliminary version, and with this behavior the model converges SEMI-CLOSE to the SVAE current spectrum in considerably less epochs. So stay tuned on that one, the transformer did converge. The behavior itself is validated and convergent in the H2 protocol spectrum.
The transformer operates with the "single" setting.
AbstractPhil/geolip-svae-transformer
I've implanted a rigid formula that allows this direct behavior from the H2 battery to superimpose onto adjacent structural boundaries, and with that built aleph and void into the system as well. These are guarantees.
As for the centrifuge concept. The optimization on the centrifuge was quite lackluster. The hardware doesn't support such behavior. You can access the current operating version of the centrifuge by utilizing "stacked" configuration. Four lenses was too much when running a quaternion bank to handle such complex interactions reasonably, so I will need to work something out in the future to get a full centrifuge system working.
Crusher is ready, transformer_v3.
You might be curious WHY these converge at such low raw MSE in the later stages. The reasoning is kind of difficult to explain, so I'll try to make it simple. The direction is very subtle in the later stages of training with AdamW, so the curves start to create much more accurate shifts towards the goals. This allows the model to rapidly converge after earlier heavier training. You can't simply train it low, it takes too long. This allows the model to KIND OF get everything NEAR where it's supposed to be, which allows the really small twitches of MSE to provide massive corrections without needing hard logits or more difficult to finetune features.
The transformer operates with the "single" setting.
AbstractPhil/geolip-svae-transformer
I've implanted a rigid formula that allows this direct behavior from the H2 battery to superimpose onto adjacent structural boundaries, and with that built aleph and void into the system as well. These are guarantees.
As for the centrifuge concept. The optimization on the centrifuge was quite lackluster. The hardware doesn't support such behavior. You can access the current operating version of the centrifuge by utilizing "stacked" configuration. Four lenses was too much when running a quaternion bank to handle such complex interactions reasonably, so I will need to work something out in the future to get a full centrifuge system working.
Crusher is ready, transformer_v3.
You might be curious WHY these converge at such low raw MSE in the later stages. The reasoning is kind of difficult to explain, so I'll try to make it simple. The direction is very subtle in the later stages of training with AdamW, so the curves start to create much more accurate shifts towards the goals. This allows the model to rapidly converge after earlier heavier training. You can't simply train it low, it takes too long. This allows the model to KIND OF get everything NEAR where it's supposed to be, which allows the really small twitches of MSE to provide massive corrections without needing hard logits or more difficult to finetune features.
- 9 replies
prithivMLmods
posted an update 2 months ago
Post
5620
I've made 8 Spaces in the Qwen-Image-Edit series, and out of them, 5 Spaces reached “Space of the Week”! A few Spaces are still topping the list even after many months.
Cumulatively, the series has crossed 8.2 million+ ZeroGPU runs and nearly 4 million visitors overall.
Thanks for all the community support! 🤗❤️
🔗 Spaces: https://huggingface.co/collections/prithivMLmods/image-generation-apps-collection
Cumulatively, the series has crossed 8.2 million+ ZeroGPU runs and nearly 4 million visitors overall.
Thanks for all the community support! 🤗❤️
🔗 Spaces: https://huggingface.co/collections/prithivMLmods/image-generation-apps-collection
- 4 replies