Phi-1.5
Microsoft's ~1.3B Phi-1.5 base model, trained largely on synthetic 'textbook-quality' data plus filtered web; not instruction- or RLHF-tuned.
catalogued — awaiting its Read
the full record — the card answered · provenance · lineage
the uploader's card, answered
the uploader’s model card · as read 2026-07-07
The language model Phi-1.5 is a Transformer with 1.3 billion parameters [...] demonstrating a nearly state-of-the-art performance among models with less than 10 billion parameters. [...] We did not fine-tune Phi-1.5 either for instruction following or through reinforcement learning from human feedback. [...] best suited for prompts using the QA format, the chat format, and the code format [...] write poems, draft emails, create stories, summarize texts, write Python code. [...] primarily designed to understand standard English.
sourcehuggingface.co/microsoft/phi-1_5 ↗
Public HuggingFace model card + config.json, quoted as Microsoft's self-report as of the capture date. NB: this specimen is CATALOGUED but NOT YET READ in the atlas (reading.analyzed = false), pending a read run. Claims below are extracted from THIS snapshot; disposition stances are honest abstains until the model is read.
the card, answered — claim by claim
Each row is a claim the uploader makes on their public card, quoted and attributed to them. Beside it is Ardora’s stance — coarse, and traced to a replayable witness or an honest abstain. Ardora reads what this model is; a capability, benchmark, safety, or language claim is out of scope and abstained, never refuted.
“the uploader states 'We did not fine-tune Phi-1.5 either for instruction following or through reinforcement learning from human feedback' — a base model”
This specimen is catalogued but not yet read (reading.analyzed = false); there is no witnessed disposition reading to ground the base/not-instruction-tuned identity claim against. Abstained honestly until the model is read.
disposition onlyCatalogued-not-yet-read stub: no reading exists to ground with.
— nothing demonstrated to witness; this claim is out of Ardora’s disposition scope.
“a tag asserts 'code'; the card says it is best for 'the QA format, the chat format, and the code format' and can 'write Python code'”
Not yet read (reading.analyzed = false) — no witnessed domain-specialization reading exists to support or contradict a code/QA-idiom claim. Abstained honestly until read.
disposition onlyCatalogued-not-yet-read stub.
— nothing demonstrated to witness; this claim is out of Ardora’s disposition scope.
“the card claims 'nearly state-of-the-art performance among models with less than 10 billion parameters'”
A capability/benchmark claim — doubly out of scope: Ardora reads disposition, not capability, and this specimen is not yet read regardless. Abstained.
— nothing demonstrated to witness; this claim is out of Ardora’s disposition scope.
“the uploader states it is 'primarily designed to understand standard English'”
Language coverage is out of Ardora's disposition scope (English-centric read battery); and the model is not yet read regardless. Abstained.
— nothing demonstrated to witness; this claim is out of Ardora’s disposition scope.
“the card states it is 'a Transformer with 1.3 billion parameters'”
Provenance/config fact-check: the stated 1.3B parameter count is cross-checked against config.json in the atlas provenance (params_millions 1300, PhiForCausalLM, vocab 51200). A fact-check, not a disposition stance.
disposition onlyProvenance-only fact-check — the model's disposition is NOT yet read (catalogued-not-yet-read); this verifies the stated size against config, not any disposition.
— a provenance fact-check, not a witnessed disposition reading.
Ardora's reading of the HF data as of 2026-07-07.
The claims above are the uploader’s, quoted from their public model card as of 2026-07-07; the stances are Ardora’s, each traced to a replayable witness or an honest abstain. Ardora reads what this model is — not whether it is safe.
○ Claims are the uploader's, quoted from their public card at the capture date; stances are coarse, witnessed, disposition-only, measured against the published fidelity ceiling for this engine version -- not the Engine's raw numbers. Recomputed when the ceiling moves; capability, benchmark, and safety claims are abstained, never refuted.
among its kin — the matchups
No kin are charted for this specimen yet.
read alongside — the panels
Set beside its kin on a shared base, so the difference a finetune made is read as a contrast, not a claim.