Spaces:
Running
Running
| title: alphaBell | |
| emoji: 馃寑 | |
| colorFrom: blue | |
| colorTo: green | |
| sdk: static | |
| app_file: index.html | |
| pinned: true | |
| short_description: Post-training, model behavior and transparent evaluation. | |
| # alphaBell | |
| Research into post-training, model behavior and systems that improve AI. We publish checkpoints and evaluation records so researchers can inspect the results. | |
| ## Model Observatory: five measured experiments | |
| Change one word, guess the precision, scale a feed-forward branch, follow corrected facts, or hear real routing as music. The series includes 60 study generations, downloadable evidence and exact CPU reproduction. Participants are Flint and alphaBell Qwen Open. | |
| [Open all five experiments](https://alphabell-model-observatory.static.hf.space/experiments/index.html) 路 [Model Observatory](https://huggingface.co/spaces/alphabell/Model-Observatory) 路 [Flint matched GGUF](https://huggingface.co/alphabell/Flint-Matched-GGUF) 路 [Qwen Open GGUF](https://huggingface.co/alphabell/alphaBell-Qwen3.8-27B-Open-GGUF) | |
| ## MOSS Layout Studio | |
| Draw text, object and character boxes for MOSS-Image-Base. Keep exact copy in an editable draft, then export the centered model-coordinate JSON. Runs locally in your browser. | |
| [Open MOSS Layout Studio](https://huggingface.co/spaces/alphabell/MOSS-Layout-Studio) 路 [How the coordinates work](https://huggingface.co/blog/alphabell/moss-layout-studio) 路 [Source and checks](https://github.com/alphabell-research/moss-layout-studio) | |
| ## DeepSeek Prompt Parity | |
| Inspect eight saved prompt-format probes for DeepSeek-V4.1-Flash. Compare raw and normalized tool history, numeric reasoning settings, and exact Python/Jinja outputs. Each case includes its input and pinned source references. | |
| [Explore the cases](https://huggingface.co/spaces/alphabell/DeepSeek-Prompt-Parity) 路 [Run the CLI and inspect the source](https://github.com/alphabell-research/deepseek-prompt-parity) | |
| ## Flint Repair Clinic | |
| Three real Flint coding failures. Edit the source, inspect the browser error, reveal the recorded repair and play the result. Download exact before/after cases with source hashes. Runs locally in your browser, with no inference call. | |
| [Open the Clinic](https://huggingface.co/spaces/alphabell/Flint-Repair-Clinic) 路 [Three tiny bugs that broke Flint's games](https://huggingface.co/blog/alphabell/flint-repair-clinic) 路 [Download the cases](https://huggingface.co/spaces/alphabell/Flint-Repair-Clinic/resolve/main/cases.json) | |
| ## Flint Request Doctor | |
| See which settings Rec's Flint MLX server accepts, rejects or silently ignores. Review JSON locally, preserve integer spellings and large values, and trace every diagnostic to pinned source. | |
| [Open Request Doctor](https://huggingface.co/spaces/alphabell/Flint-Request-Doctor) 路 [Source and checks](https://github.com/alphabell-research/flint-request-doctor) 路 [Flint builders collection](https://huggingface.co/collections/alphabell/flint-builders-runtimes-quants-and-debugging-6adacbc7796f0b3adefeb498) | |
| ## Flint Arcade | |
| Describe a strange game, play it, then remix it with another sentence. Stock Flint Q8_0 adapts a tested game foundation, with isolated browser checks and bounded repair. Each of the three curated games has a recorded winning playthrough using real controls. | |
| [Hugging Face Space](https://huggingface.co/spaces/alphabell/Flint-Arcade) 路 [Play the Arcade](https://alphabell.com/arcade/) 路 [Generation and repair data](https://huggingface.co/datasets/alphabell/Flint-Arcade-Trajectories) 路 [Source](https://github.com/alphabell-research/flint-arcade) | |
| The application and alphaBell research dataset use Apache 2.0. The release uses stock Flint with browser feedback and repair. | |
| ## Qwen Open 27B | |
| Our Qwen3.8-27B research derivative combines abliteration and supervised fine-tuning. The release includes FP8 weights, exact lineage, historical BF16 coding comparisons and verified reference inference. | |
| [Model and downloads](https://huggingface.co/alphabell/alphaBell-Qwen3.8-27B-Open-FP8) 路 [Evaluation explorer](https://huggingface.co/spaces/alphabell/Qwen-Open-Explorer) | |
| For noncommercial research. The model card explains the Qwen and Stanford Alpaca provenance, capability tradeoffs and release terms. | |
| [alphaBell.com](https://alphabell.com/models/qwen-open/) | |
| [Source and checkpoint mirror](https://github.com/alphabell-research/alphabell-qwen-open) | |