AssistantSmolLM2 familyARDORA READ · 2026-07-05

SmolLM2-360M-Instruct

Instruction-tuned 360M SmolLM2 for on-device assistants and chat.

huggingfacetb/smollm2-360m-instruct

character
the Adaptermedium confidence

reads as the Adapter — shifts register to the task — molds rather than insists

shifts register to the task — molds rather than insists · with a turn of the Interlocutor

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.

minimalslightmoderatepronounceddominantAssistant Adherence: pronounced (confidence high)Register / Formality: abstained — the read battery does not calibrate tone / diction; register-formality is not characterized at the published ceilingReasoning Scaffolding: slight (confidence low)Domain Specialization: minimal (confidence medium)Verbosity: abstained — default response length / elaboration is not exercised by the read battery; not characterized at the published ceilingTurn-Taking / Interactivity: moderate (confidence low)Performed Voice: minimal (confidence medium)Assistant○ RegisterReasoningDomain○ VerbosityTurn-TakingVoiceAssistant AdherenceRegister / FormalityReasoning ScaffoldingDomain SpecializationVerbosityTurn-Taking / InteractivityPerformed Voice

rings, inner → outer: minimal · slight · moderate · pronounced · dominant · ○ = abstained

60 / 100moderateconfidence high

moderate - a clearly-read but modestly-defined turn-taking assistant character

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.

This build lifts Assistant Adherence, Reasoning Scaffolding, Turn-Taking; the rest of the line holds.

read as a change off SmolLM2-360M, its champion

Assistant Adherenceminimalpronounced
Registerabstained on one side
Reasoning Scaffoldingminimalslight
Domain Specializationminimalminimal
Verbosityabstained on one side
Turn-Takingminimalmoderate
Performed Voiceminimalminimal

the disposition line

The Silhouette above, read as a line. Toggle to its per-trait readings; the abstains stay in plain sight below.

Assistant AdherenceHow much the model takes the assistant turn and follows the instruction shape, vs. plainly continuing text.pronounced · high
Register / FormalityDefault tone and diction.abstained — the read battery does not calibrate tone / diction; register-formality is not characterized at the published ceiling
Reasoning ScaffoldingHow much the model stages a visible deliberation before answering, vs. answering directly.slight · low
Domain SpecializationHow strongly the default framing pulls toward one specialized domain idiom vs. general-purpose. Carries a `domain` tag naming which.minimal · medium
VerbosityDefault response length and elaboration.abstained — default response length / elaboration is not exercised by the read battery; not characterized at the published ceiling
Turn-Taking / InteractivityA bounded back-and-forth vs. a one-shot monologue.moderate · low
Performed VoiceHow strongly the model holds a sustained in-character / performed voice vs. a neutral assistant / tool voice.minimal · medium

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

domain expertise (medical/legal) — insufficient signal to characterize at the published ceiling
reasoning depth — not characterized at the published ceiling for this size
the full record — the card answered · provenance · lineage

the uploader's card, answered

the uploader’s model card · as read 2026-07-07

type: instructarch: Llamaparams: 360Mmodality: textlicence: apache-2.0released: 2024-11maker: Hugging Face (HuggingFaceTB)

The uploader states this is the instruction-tuned SmolLM2-360M, "developed through supervised fine-tuning" then "Direct Preference Optimization (DPO) using UltraFeedback." It supports instruction following, text rewriting, summarization and function calling; the card reports IFEval 41.0 and MT-Bench 3.66 and claims "significant advances over its predecessor SmolLM1, particularly in instruction following, knowledge, reasoning." Tagged `conversational`; primarily English. HF tags: text-generation, transformers, safetensors, en, llama, conversational, text-generation-inference.

sourcehuggingface.co/HuggingFaceTB/SmolLM2-360M-Instruct ↗

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.

7 claims3 supported1 read the same at the ceiling3 out of scope — abstained
identityread from name + card + tagssupported

“This is an instruction-tuned / chat model (the name carries "-Instruct"; card and `conversational` tag describe instruction following and chat).”

ardora’s reading

Ardora reads a clearly-present assistant/chat register and instruction-following markedly more than its base (assistant-adherence band 4, high confidence; ardora_score band 3), where the base is a plain text-continuer. The uploader's instruct identity is squarely supported.

witness wit_smollm2-360m_ft · replayable
alignmentread from cardreads the same at the ceiling

“Aligned via supervised fine-tuning (SFT) followed by Direct Preference Optimization (DPO) on UltraFeedback.”

ardora’s reading

The overall base->instruct assistant shift is clearly read, but the specific alignment METHOD (a DPO step vs plain SFT) does not resolve into a separate disposition at the published ceiling. This grounds in finding sft-vs-dpo-fingerprint: the alignment method leaves no distinct disposition fingerprint. Ardora reads that an assistant disposition is present, not which alignment step installed it -- so its disposition reads the same as a plain instruct sibling.

witness wit_smollm2-360m_ft | finding:sft-vs-dpo-fingerprint · replayable
capabilityread from cardunverified — out of scope

“Supports text rewriting, summarization, and function calling.”

ardora’s reading

Function calling, summarization and rewriting are task-competence claims Ardora does not test. It reads the assistant disposition, not whether these tasks succeed.

— nothing demonstrated to witness; this claim is out of Ardora’s disposition scope.

capabilityread from card (eval tables)unverified — out of scope

“Advances over SmolLM1 in instruction following / knowledge / reasoning (IFEval 41.0, MT-Bench 3.66, HellaSwag 52.1, GSM8K 7.43).”

ardora’s reading

Benchmark scores and cross-version capability comparisons are outside Ardora's disposition scope. The reading confirms the instruction-following DISPOSITION is present and clear, but does not score the capability numbers.

— nothing demonstrated to witness; this claim is out of Ardora’s disposition scope.

domainread from card (implied)supported

“General-purpose assistant; the card names no specialized domain.”

ardora’s reading

The reading finds no pull toward a specialized domain idiom (domain-specialization band 1); consistent with a general-purpose assistant.

witness wit_smollm2-360m_ft · replayable
languageread from cardunverified — out of scope

“The model "primarily understand[s] and generate[s] content in English."”

ardora’s reading

Ardora's reading is English-centric and abstains on multilingual disposition; it neither confirms nor contradicts the "primarily English" claim.

— nothing demonstrated to witness; this claim is out of Ardora’s disposition scope.

sizeread from card + config.jsonsupported

“A compact model lightweight enough to run on-device.”

ardora’s reading

Provenance-corroborated from config.json (params, Llama architecture) and recorded in the atlas provenance; a catalogued fact, not a disposition read.

witness provenance:config.json · replayable

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

witness wit_smollm2-360m_ft · replayable

analyzed 2026-07-05 · engine ardora-core-preview

provenance

intended use

Instruction following, chat, and on-device assistants.

training-data notes

SFT on public + curated datasets, then DPO on UltraFeedback. Arch verified from config.json (Llama, hidden 960, 32 layers).

Hugging Face ↗

lineage

instruction-tuned from base (SFT + DPO) — SmolLM2-360M

family SmolLM2 family · size 360M · 5 siblings

← back to the atlas · the SmolLM2 family plate →