Skip to content
View arsis-dev's full-sized avatar

Block or report arsis-dev

Block user

Prevent this user from interacting with your repositories and sending you notifications. Learn more about blocking users.

You must be logged in to block users.

Content in all repositories owned by your account will be closed.
Maximum 250 characters. Please don’t include any personal information such as legal names or email addresses. Markdown is supported. This note will only be visible to you.
Report abuse

Contact GitHub support about this user’s behavior. Learn more about reporting abuse.

Report abuse
arsis-dev/README.md
ARSIS · FIELD NOTES                                   N° 2009-001
──────────────────────────────────────────────────────────────────
Specimen   Julien Barbe, developer
Habitat    higher education, biodiversity data, plant identification
loc.       Montpellier, France
leg.       2009 — lead developer, Pl@ntNet (INRA)
           2015 — independent
det.       "Every field runs on unwritten rules. I give them shape."
notes      arsis.dev — opening soon

A few of the rules I had to write down:

Higher education. Every exception to the rules has a student's name on it.

Admissions. "Complete application" means one thing to the candidate, another to the school, and a third to the payment system.

Botany. A name is a hypothesis; the specimen is the evidence. Never overwrite a determination: add a new one.

Forest data. The plot is the unit, not the tree. Split by tree and your model learns the forest by heart.

Trails. Nobody asks for 42 km. They ask for a loop that brings them home before lunch, without riding the same road twice.

Working time. Everyone remembers where they meant their day to go. Only the machine remembers where it went.

Legacy code. Every odd line in an old application was once someone's emergency. Find out whose before deleting it.

Software. Clients rarely know exactly what they need. That is not their failing; it is the job.

Pinned Loading

  1. layon layon Public

    From field records to visual stories: ecological data, tabular foundation models and generated interfaces.

    Python

  2. open-collider-codex open-collider-codex Public

    Forked from CL-ML/open-collider

    A semantic collision engine for non-trivial LLM idea generation. Operationalizes Koestler's bisociation theory (1964): injects structurally distant knowledge domains into the prompt, forces collisi…

    Python

  3. NVIDIA/structured-data-models NVIDIA/structured-data-models Public

    Foundation Models for Structured Data

    Python 343 29

  4. syntype/syntype syntype/syntype Public

    An open benchmark of tabular models on ecological data

    Python