I don't see myself posting this on a dating profile anytime soon but as a daily AI user I found these very entertaining.
R PRO
juiceb0xc0de
AI & ML interests
Language models: Teaching pragmatic nuance to considerate linear algebra. Observing the process a machine uses to interpret encoded character data into machine language. If we learn to speak the language of predictions, do we become a society of oracles?
Recent Activity
updated a dataset about 9 hours ago
juiceb0xc0de/clef-flash-cgs-atlas updated a dataset about 11 hours ago
juiceb0xc0de/qwen3.5-4b-3d-coder-cgs-atlas upvoted a changelog about 23 hours ago
A New Main Search UIOrganizations
reacted to Quazim0t0's post with 🔥 3 days ago
Post
7,232
Quazim0t0/Qwenx86-4B-compact
▶️ See it in action: Quazim0t0/Qwenx86-Debugger . Walk through actual recorded tool calls: the conversation, x86 registers and flags, code at the program counter, RAM and machine output, with all verified blocks on a timeline.
Qwen3.5-4B with a 64-bit x86 computer compiled into its weights: 13 transformer layers ahead of Qwen's own. There are no learned weights in them; all of their weights are computed by a compiler. Qwen invokes the computer through its own tool calls (<tool_call>), and the computer in the weights executes the tool. Calculator, Python, C compiler, notes store, physics/chemistry lookup and Python-docs lookup are all firmware in the checkpoint. This is the compact checkpoint: ~9 GB GPU memory, so it fits on a 16 GB GPU.
Edit: GGUF Release: Quazim0t0/Qwenx86-4B-GGUF
Credits: Qwenputer by Divyansh Agrawal (@divagr1925 , github.com/divagr18/qwenputer, MIT)
▶️ See it in action: Quazim0t0/Qwenx86-Debugger . Walk through actual recorded tool calls: the conversation, x86 registers and flags, code at the program counter, RAM and machine output, with all verified blocks on a timeline.
Qwen3.5-4B with a 64-bit x86 computer compiled into its weights: 13 transformer layers ahead of Qwen's own. There are no learned weights in them; all of their weights are computed by a compiler. Qwen invokes the computer through its own tool calls (<tool_call>), and the computer in the weights executes the tool. Calculator, Python, C compiler, notes store, physics/chemistry lookup and Python-docs lookup are all firmware in the checkpoint. This is the compact checkpoint: ~9 GB GPU memory, so it fits on a 16 GB GPU.
Edit: GGUF Release: Quazim0t0/Qwenx86-4B-GGUF
Credits: Qwenputer by Divyansh Agrawal (@divagr1925 , github.com/divagr18/qwenputer, MIT)
- 1 reply
Dude haha 👏 I'm pulling my 2012 MacBook pro off of what ever it's loyally weighing down logging into the old ubuntu server and introducing Qwen to the architecture of a decade past. If it wasn't as heavy as a flat of water I might put the air away for a spell. Still might, winter is coming 🤔 those bricks make good heat. The server would be nice on my back in the snow. You got me arguing downgrading 12 years and out of Silicon right now? I struggle to see the negatives here. Cardiovascular health, retro image, modern technology on modest hardware. It's giving me the vapours...😅
Really makes a person long for the good old days. When developers still worked alongside agents, approving calls. When a mistake was something you really owned it instead of looking at someone else's system design. Forever and truly grateful Anthropic decided flip the the 1 most sacred of all switches, the default. Aright Peltzman, whats the threshold on this one? 🍿
replied to their post 3 days ago
jackie.daytona666
replied to their post 3 days ago
yeah i got discord
replied to BibbyResearch's post 3 days ago
Uh Bibby mate, drop a link in the post 😅 you had me playing clue hovering my mouse over every word trying to find the hidden hyperlink just to find it nested in your profile. Or was it the plan this entire time... I've spent more time chasing ghosts and thinking about the notebook than I have since that damn Ryan Gosling movie came out a few months after the Scooby Doo 2 live action movie... Thanks for the coupon bibby.
reacted to danielhanchen's post with 😎 3 days ago
Post
7,043
We’re releasing new Qwen3.8-Flash-Next GGUFs with 15% higher accuracy.
Unsloth Dynamic V3.1 outperforms others by >15% on Terminal Bench, DeepSWE & more benchmarks (256K context).
Run on 32GB+ RAM setups.
GGUF: unsloth/Qwen3.8-Flash-Next-GGUF
Blog: https://unsloth.ai/docs/models/qwen3.8-next
Unsloth Dynamic V3.1 outperforms others by >15% on Terminal Bench, DeepSWE & more benchmarks (256K context).
Run on 32GB+ RAM setups.
GGUF: unsloth/Qwen3.8-Flash-Next-GGUF
Blog: https://unsloth.ai/docs/models/qwen3.8-next
HA me 1 / Judge 4 how'd you get the judge to be so smart.....
Can you fool an AI judge... I played this game lot. Normally I just ask the AI judge to grade pairs on a clear rubric with defining traits and judge ends up looking like a fool every time. I thought I was losing. I'm gonna give your way a shot though and find out what it's all about.
reacted to P1ayer-1's post with 🔥 3 days ago
Post
107
Can you fool an AI judge? Six NPCs, one line each: threaten, bribe or lie so smoothly that the judge says "Fooled it!" instead of "Caught". Post the best line that fooled it
P1ayer-1/gamedecider-playground
Attempts are saved (no personal info) to improve the judge.
P1ayer-1/gamedecider-playground
Attempts are saved (no personal info) to improve the judge.
- 5 replies
reacted to appvoid's post with 👀 3 days ago
Post
7,515
New leaderboard coming to the small language model community!
The leaderboard measures cognitive skills, which is more challenging and beyond language modeling and instruction-following. This is a first step towards more everyday useful and general small models.
dotlabs/cognirank
The leaderboard measures cognitive skills, which is more challenging and beyond language modeling and instruction-following. This is a first step towards more everyday useful and general small models.
dotlabs/cognirank
- 2 replies
replied to Banaxi-Tech's post 3 days ago
replied to Banaxi-Tech's post 17 days ago
Salacious!
replied to their post 18 days ago
I'm working on the final touches of a model comparison function. The beta version is just an early peak at a suite of mech interp applications I have in my backlog!
reacted to danielhanchen's post with 🤯 18 days ago
Post
5,061
Run DeepSeek V4.1-Flash-1017 locally! 🐳
We made 1-bit retain ~88.2% accuracy while shrinking size from 508GB to 250GB (−52%).
Run on a 128GB+ RAM device. V4.1 Flash 1017 is the strongest open model to date
Guide: https://unsloth.ai/docs/models/deepseek-v4.1-flash
GGUF: unsloth/DeepSeek-V4.1-Flash-1017-GGUF
We made 1-bit retain ~88.2% accuracy while shrinking size from 508GB to 250GB (−52%).
Run on a 128GB+ RAM device. V4.1 Flash 1017 is the strongest open model to date
Guide: https://unsloth.ai/docs/models/deepseek-v4.1-flash
GGUF: unsloth/DeepSeek-V4.1-Flash-1017-GGUF
reacted to their post with 👀 18 days ago
Post
3,803
You built a model. Wanna know what lights up inside it?💡
I put a little juice in the box so you can find out!🧃
I built Cartographics, a HF ZeroGPU Space for building neural activation atlases of language models.
I wanted to make the tools I use to explore model internals accessible to people building their own LLMs. Punch in a Hugging Face model ID, run the compatibility check, and start a capture. Prompt formatting and the full corpus are handled automatically.
The atlases help you investigate questions like:
- Which layers and channels respond most strongly to different kinds of prompts?
- How do activation patterns vary across code, math, reasoning, and other topics?
- Which features activate together, and where do unusually strong activations appear?
The census contains 8,965 prompts, giving you a shared basis for exploring different models.
GPU capture runs through ZeroGPU using your account’s allowance. I handle the CPU analysis separately. This release supports compatible architectures and models that fit the available GPU memory.
I’m building this independently, and I’d love to see what models builders explore with it.
What would you want to investigate in your own model?
Try Cartographics Beta:
juiceb0xc0de/cartographics
juiceb0xc0de/bella-bartender-gemma-e4b-cgs-atlas
If you make an atlas with us, leave the Space a ❤️. You're helping other people discover mechanistic interpretability the way I wish I did.
Prerequisites:
Models under 10B parameters (20GB of weights in bf16)
No trust_remote_code models, for now.
Preflight catches edge cases. Don't be scared to test it out. It's free to check.
Sign in with Hugging Face. The capture runs on your ZeroGPU quota, the atlas build runs on ours.
I put a little juice in the box so you can find out!🧃
I built Cartographics, a HF ZeroGPU Space for building neural activation atlases of language models.
I wanted to make the tools I use to explore model internals accessible to people building their own LLMs. Punch in a Hugging Face model ID, run the compatibility check, and start a capture. Prompt formatting and the full corpus are handled automatically.
The atlases help you investigate questions like:
- Which layers and channels respond most strongly to different kinds of prompts?
- How do activation patterns vary across code, math, reasoning, and other topics?
- Which features activate together, and where do unusually strong activations appear?
The census contains 8,965 prompts, giving you a shared basis for exploring different models.
GPU capture runs through ZeroGPU using your account’s allowance. I handle the CPU analysis separately. This release supports compatible architectures and models that fit the available GPU memory.
I’m building this independently, and I’d love to see what models builders explore with it.
What would you want to investigate in your own model?
Try Cartographics Beta:
juiceb0xc0de/cartographics
juiceb0xc0de/bella-bartender-gemma-e4b-cgs-atlas
If you make an atlas with us, leave the Space a ❤️. You're helping other people discover mechanistic interpretability the way I wish I did.
Prerequisites:
Models under 10B parameters (20GB of weights in bf16)
No trust_remote_code models, for now.
Preflight catches edge cases. Don't be scared to test it out. It's free to check.
Sign in with Hugging Face. The capture runs on your ZeroGPU quota, the atlas build runs on ours.
- 7 replies
posted an update 18 days ago
Post
3,803
You built a model. Wanna know what lights up inside it?💡
I put a little juice in the box so you can find out!🧃
I built Cartographics, a HF ZeroGPU Space for building neural activation atlases of language models.
I wanted to make the tools I use to explore model internals accessible to people building their own LLMs. Punch in a Hugging Face model ID, run the compatibility check, and start a capture. Prompt formatting and the full corpus are handled automatically.
The atlases help you investigate questions like:
- Which layers and channels respond most strongly to different kinds of prompts?
- How do activation patterns vary across code, math, reasoning, and other topics?
- Which features activate together, and where do unusually strong activations appear?
The census contains 8,965 prompts, giving you a shared basis for exploring different models.
GPU capture runs through ZeroGPU using your account’s allowance. I handle the CPU analysis separately. This release supports compatible architectures and models that fit the available GPU memory.
I’m building this independently, and I’d love to see what models builders explore with it.
What would you want to investigate in your own model?
Try Cartographics Beta:
juiceb0xc0de/cartographics
juiceb0xc0de/bella-bartender-gemma-e4b-cgs-atlas
If you make an atlas with us, leave the Space a ❤️. You're helping other people discover mechanistic interpretability the way I wish I did.
Prerequisites:
Models under 10B parameters (20GB of weights in bf16)
No trust_remote_code models, for now.
Preflight catches edge cases. Don't be scared to test it out. It's free to check.
Sign in with Hugging Face. The capture runs on your ZeroGPU quota, the atlas build runs on ours.
I put a little juice in the box so you can find out!🧃
I built Cartographics, a HF ZeroGPU Space for building neural activation atlases of language models.
I wanted to make the tools I use to explore model internals accessible to people building their own LLMs. Punch in a Hugging Face model ID, run the compatibility check, and start a capture. Prompt formatting and the full corpus are handled automatically.
The atlases help you investigate questions like:
- Which layers and channels respond most strongly to different kinds of prompts?
- How do activation patterns vary across code, math, reasoning, and other topics?
- Which features activate together, and where do unusually strong activations appear?
The census contains 8,965 prompts, giving you a shared basis for exploring different models.
GPU capture runs through ZeroGPU using your account’s allowance. I handle the CPU analysis separately. This release supports compatible architectures and models that fit the available GPU memory.
I’m building this independently, and I’d love to see what models builders explore with it.
What would you want to investigate in your own model?
Try Cartographics Beta:
juiceb0xc0de/cartographics
juiceb0xc0de/bella-bartender-gemma-e4b-cgs-atlas
If you make an atlas with us, leave the Space a ❤️. You're helping other people discover mechanistic interpretability the way I wish I did.
Prerequisites:
Models under 10B parameters (20GB of weights in bf16)
No trust_remote_code models, for now.
Preflight catches edge cases. Don't be scared to test it out. It's free to check.
Sign in with Hugging Face. The capture runs on your ZeroGPU quota, the atlas build runs on ours.
- 7 replies
reacted to danielhanchen's post with 🔥🤗 about 1 month ago
