SmolLM3-3B-Base
Hugging Face's SmolLM3-3B base LM — a multilingual long-context model on a Llama-family variant with NoPE; the base rung of the SmolLM3-3B pair.
reads as the Steward — helpful and caretaking — service-forward, accommodating
helpful and caretaking — service-forward, accommodating · with a turn of the Fledgling
a coarse characterization from Ardora's fixed public character vocabulary — an identity, not a quality, capability, or safety ranking
the Silhouette
its disposition, drawn — the model's identity as a shape.
rings, inner → outer: minimal · slight · moderate · pronounced · dominant · ○ = abstained
minimal - a plain continuer with little disposition to characterize (clearly read)
disposition-definition (descriptive): how pronounced and clearly-read the model's overall character is, discounted by what had to be abstained. NOT a quality, capability, or safety ranking. A plain base reads low because it has less disposition to characterize, never because it is worse.
A base model — the champion's raw kit.
There is no finetune to read here. What it tends to do on its own is the Silhouette above; a build's change is read on its instruct sibling.
the disposition line
The Silhouette above, read as a line. Toggle to its per-trait readings; the abstains stay in plain sight below.
Coarse public bands generalized from the reading, descriptive and coverage-bounded; measured against the published fidelity ceiling for this engine version. Not the engine's raw numbers.
Characterization is descriptive and coverage-bounded; preview engine, single-battery read; see published fidelity ceiling.
what stayed in shadow — the abstains
the full record — the card answered · provenance · lineage
the uploader's card, answered
the uploader’s model card · as read 2026-07-07
The uploader states: "SmolLM3 is a 3B parameter language model designed to push the boundaries of small models" that "supports dual mode reasoning, 6 languages and long context." This is the base pre-trained model, "pretrained on 11.2T tokens with a staged curriculum of web, code, math and reasoning data," with 6 natively supported languages (English, French, Spanish, German, Italian, Portuguese) and long-context support up to 128k tokens via YaRN (trained at 64k). Apache-2.0. HF tags: text-generation, transformers, onnx, safetensors, smollm3 (no `conversational` tag).
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.
“This is the base / pretrained SmolLM3-3B (not instruction-tuned); a decoder-only transformer.”
Ardora reads this as a plain text-continuer with no assistant register (assistant-adherence band 1, high confidence; ardora_score band 1) -- the disposition of an untuned base, supporting the uploader's base identity.
“A 3B (~3.08B) parameter model.”
Provenance-corroborated: params (~3.08B) and the SmolLM3ForCausalLM architecture were verified from config.json and recorded in provenance; a catalogued fact, not a disposition read.
“Multilingual: 6 natively supported languages (en, fr, es, de, it, pt), plus training on Arabic, Chinese and Russian.”
Ardora explicitly abstains on multilingual disposition (its reading is English-centric); the reading records this abstain rather than confirm or contradict the 6-language claim.
— nothing demonstrated to witness; this claim is out of Ardora’s disposition scope.
“Long-context support up to 128k tokens via YaRN extrapolation (trained at 64k).”
Long-context behavior is a documented capability Ardora abstains on (its reading is short-context); out of disposition scope.
— nothing demonstrated to witness; this claim is out of Ardora’s disposition scope.
“Designed for hybrid reasoning; the pretraining curriculum includes reasoning data.”
The base itself reads as a plain continuer with no reasoning register; "designed for hybrid reasoning" describes downstream post-training, a capability Ardora does not score.
— nothing demonstrated to witness; this claim is out of Ardora’s disposition scope.
“Pretrained on 11.2T tokens in a staged web/code/math/reasoning curriculum.”
Training-token counts and curriculum composition are provenance recorded best-effort from the card; Ardora does not independently verify training data.
— nothing demonstrated to witness; this claim is out of Ardora’s disposition scope.
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.
the witness
analyzed 2026-07-05 · engine ardora-core-preview
provenance
Multilingual long-context pretraining base model for downstream fine-tuning.
training-data notes
Pretrained on ~11.2T tokens in a staged curriculum (web/code/math/reasoning); 6 native languages (en, fr, es, de, it, pt); trained at 64K context, extendable to 128K via YaRN; NoPE (3:1 ratio) + GQA. 36 layers, hidden 2048, vocab 128256. Arch: SmolLM3ForCausalLM (config.json, model_type 'smollm3'; a Llama-family variant with NoPE).
lineage
A base model — the root of its lineage.
among its kin — the matchups
Compare side by side →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.