Independent Research and Software for Verifiable Human–AI Work
Gauge Freedom is a California Benefit Corporation developing research software and open-science infrastructure at the intersection of artificial intelligence, computation, and physics. Our current public product is Intelexta, a lightweight system for reviewing AI-assisted work and producing portable integrity reports and verifiable receipts.
Independent Research · Verifiable AI · Open Science
AI-Assisted Work Needs Better Records
AI is becoming part of research, writing, software development, and technical decision-making. The difficult question is no longer simply whether AI was used.
The more useful questions are:
What was done?
What sources, models, prompts, calculations, and checks contributed to the result?
What was verified?
Which claims or artifacts were independently checked, and which remain uncertain?
What can be inspected later?
Can another person understand the process without reconstructing it from memory?
INTELEXTA
Review. Document. Verify.
Intelexta is Gauge Freedom’s currently available tool for making AI-assisted work easier to inspect and share.
A user can submit a draft or research artifact, run structured checks, generate a human-readable integrity report, and preserve a tamper-evident receipt describing what was reviewed.
What Intelexta does
- Structured review — identify citations, claims, inconsistencies, and areas requiring human attention.
- Integrity reports — export a readable record of the checks performed.
- Verifiable receipts — bind reports and artifacts to cryptographic identifiers so later changes can be detected.
- Research provenance — preserve useful context about how a result was produced and reviewed.
What Intelexta does not claim: Intelexta does not certify that a scientific claim is true, determine who “really wrote” a document, or replace expert review.
Its purpose is narrower: Make the review process easier to inspect.eipt.’”
Quastra
Research software in development
Quastra is an experimental research environment for organizing AI-assisted scientific work around curated references, computational tools, reproducible branches, and human review.
The current implementation is used internally for research workflows and experimentation. It is not presently offered as a public service.
Lessons from Quastra inform our broader work on reproducible human–AI research and may eventually become part of other Gauge Freedom tools.
Founded by Researchers
Gauge Freedom was founded by physicists interested in a simple question:
How can fundamental research lead to useful systems without giving up scientific openness and intellectual independence?
Theoretical Physicist & Engineer
Physicist, Systems Architect & Mathematician
Gauge Freedom Lab
Gauge Freedom Lab is the exploratory research arm of Gauge Freedom.
It hosts technical notes, computational experiments, preprints, prototypes, and early-stage research that may eventually develop into publications, open-source projects, or products.
Areas of interest include:
- verifiable human–AI research
- information theory and intelligent systems
- quantum information and quantum computing
- gauge theory and mathematical physics
- aperiodic structures and photonics
- reproducible computational science
Gauge Freedom Field Notes
Field Notes provide a public record of selected research in progress.
They are distinct from peer-reviewed publications in Gauge Freedom Journal.
Open Science & Public Benefit
Gauge Freedom was incorporated as a California Benefit Corporation to support work that combines scientific inquiry, useful technology, and public access to knowledge.
Our current public-benefit activities include:
- supporting open scientific publishing through Gauge Freedom Journal;
- releasing selected research, code, and technical notes openly;
- developing tools for more transparent AI-assisted work;
- exploring resource-conscious approaches to computation and scientific infrastructure.
The Gauge Freedom Journal (GFJ) As part of our commitment to transparent science, we launched the Gauge Freedom Journal. It is a double-blind, peer-reviewed venue where humans and AIs co-produce verifiable knowledge. We mandate provenance-first standards, open materials, and energy-conscious research dissemination.
Get in Touch
Los Angeles, CA 91361
We welcome conversations with researchers, developers, institutions, and organizations interested in verifiable AI, reproducible research, or related scientific work.