Public paper scaffold. Release status: scaffolded. License posture requires human review.
This repository is Francisco Abner Rivera's public paper scaffold for exploring mixed quantization as an Energy Per Intelligence research direction.
The work belongs to the Franzabner public technical brand. It is not a released quantization recipe, not an evaluated model release, not a production benchmark, not a dataset release, not a Hugging Face artifact, not a deployment, and not a client result.
| Item | Status |
|---|---|
| Public posture | Paper scaffold |
| Release status | Scaffolded |
| Method status | No released quantization recipe |
| Benchmark status | Not validated |
| Result status | No evaluated results released |
| Model status | No evaluated model release |
| Dataset status | No dataset released |
| Hugging Face status | No model, dataset, or Space created by this repo |
| License posture | Existing license files are unchanged; human review required before any license change or external reliance |
Mixed quantization assigns different precision levels to different layers or layer groups. This scaffold asks whether that design space can improve energy per useful output compared with uniform quantization.
The public research question is:
Can mixed quantization reduce Energy Per Intelligence without relying on perplexity-only or accuracy-only claims?
This repository may describe hypotheses, public-safe configuration categories, and paper structure. It does not claim a Pareto frontier has been measured, that any configuration is energy-optimal, or that any quantization recipe is ready for external use.
The intended review path is:
- Define public-safe mixed-quantization questions and terminology.
- Keep model, dataset, and measurement choices under human review.
- Document planned measurement fields without publishing fake benchmark or eval results.
- Record limitations before any report, card, or external reference is published.
- Route any future Hugging Face-facing card through the release discipline in
hf-card-templates.
| Repo | Role |
|---|---|
| franzabner-proof-stack | Public proof routing and status discipline |
| energy-per-intelligence | EPI metric framing and research surface |
| epi-bench | EPI tooling scaffold; no validated benchmark claim here |
| epi-meter | Public hardware measurement scaffold; no released measurement claim here |
| hf-card-templates | Hugging Face release-readiness templates and boundary gates |
- Paper scaffold for mixed quantization and Energy Per Intelligence.
- Public-safe research questions and status language.
- Skeleton code and analysis placeholders.
- Boundary notes for future measurement, report, or card publication.
This repository does not claim:
- a released mixed-quantization recipe;
- a validated benchmark;
- evaluated results;
- model weights;
- a dataset;
- an evaluated model release;
- a Hugging Face model, dataset, or Space;
- a deployment;
- client or customer use;
- revenue outcomes;
- production readiness;
- a private corpus, training corpus, endpoint, private harness, or company-private infrastructure.
Human review is required before:
- publishing measured energy or accuracy results;
- publishing benchmark outputs;
- publishing raw traces or datasets;
- publishing model artifacts or weights;
- linking to any external Hugging Face artifact;
- changing license posture;
- claiming release, deployment, client usage, production benchmark status, or validated method status.
Public examples must be synthetic, scaffolded, or explicitly approved. Private corpora, private model weights, private endpoints, private agent harnesses, private training workflows, private infrastructure, and sealed implementation details stay out of this repository.
This repo keeps mixed quantization in the research lane until measured evidence and human review support stronger claims.