russell@unturf.com
39f8aa1fb4
- '.. class:: center' parsed as a Python class named 'center' (duplicate
across one-pager + two-pager) -> '.. rst-class:: center' (styling, no
object). Clears the duplicate-object-description warnings.
- two v8-consensus section underlines were shorter than their titles
('underline too short') -> extended to title length.
Pager + v8 pages now build with zero warnings.
65 lines
3.2 KiB
ReStructuredText
65 lines
3.2 KiB
ReStructuredText
arborist — answers your AI can prove
|
||
=====================================
|
||
|
||
.. rst-class:: center
|
||
|
||
*one-page summary · permacomputer.com · AGPL-3.0*
|
||
|
||
----
|
||
|
||
Most retrieval-augmented question-answering systems hand a language
|
||
model some context and ship whatever the model says. There is no way to
|
||
tell whether the answer faithfully reflects the source or whether the
|
||
model embroidered it. **arborist closes that gap.** Every answer is
|
||
verified against its source *after* the language model finishes, by a
|
||
mechanical checker — not by another AI grading the first one. The
|
||
checker labels each answer **grounded** (every claim was found in the
|
||
source), **partly grounded** (some claims, not all), or
|
||
**ungrounded**. The label travels with the answer and is stored in a
|
||
tamper-evident chain back to the source bytes.
|
||
|
||
**Fabricated citations become impossible.** When the model answers, it
|
||
never types the quoted text. arborist tags each candidate source
|
||
chunk with a short label — ``E1``, ``E2`` — and asks the model to
|
||
answer using those labels. The model might write *"Jupiter is the
|
||
largest planet [E1], with a radius of about 70 000 km [E2]"*; arborist
|
||
renders the actual chunk text at display time. A model cannot fabricate
|
||
a quote it never types.
|
||
|
||
**The proof path is cheap and mechanical.** Verification is text
|
||
comparison, not embeddings, not similarity, not another model. It runs
|
||
on a laptop. Optional smart-ranking layers — cross-encoder rerankers,
|
||
entailment models — exist on the side; they help arborist find better
|
||
evidence, they do not influence whether an answer is certified.
|
||
|
||
**Same question, same document, same answer.** Answers are content-
|
||
addressed: the cache key folds in the document, the question, the
|
||
model identity, and the policy under which the answer was checked.
|
||
Two users asking the same question of the same document under the same
|
||
policy hit the same record. Reproducible. Replayable. Shareable.
|
||
|
||
**What it has measured on real traffic.**
|
||
|
||
- **100% misattribution catch at 0% false positives** — every answer
|
||
where the cited source was unrelated to the claim was flagged,
|
||
without a single grounded answer wrongly demoted, across the full
|
||
pooled test bed.
|
||
- **55–65% topic-deflection catch at 0–0.4% false positives** — picks
|
||
off-topic answers out of the stream while leaving on-topic answers
|
||
untouched.
|
||
- **100% citation coverage on the curated textbook corpus** — every
|
||
cited claim resolves to a chain of evidence ending at a public-domain
|
||
or open-licensed source.
|
||
|
||
**What it costs.** Python 3.12. SQLite, one file per shard, each
|
||
designed around ~10 GB. A Wikipedia-class corpus lands across a
|
||
handful of shards (the live deployment is four shards, ~38 GB on
|
||
disk, holding 3.5 M documents / 6.2 M chunks). No GPU for the proof
|
||
path. Use any OpenAI-compatible inference endpoint; the free
|
||
reference endpoint is `hermes.ai.unturf.com/v1
|
||
<https://hermes.ai.unturf.com/v1/models>`_ (OpenAI-compatible — the
|
||
``/v1/models`` link returns the live model card). Source:
|
||
`git.unturf.com/engineering/unturf/arborist
|
||
<https://git.unturf.com/engineering/unturf/arborist>`_. Full
|
||
whitepaper: `unfirehose.com/merkle-providence-reverse-rag.html
|
||
<https://unfirehose.com/merkle-providence-reverse-rag.html>`_.
|