© Neuronpedia 2026
    Privacy & TermsBlogGitHubSlackTwitterContact
    Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    APIAssistant AxisNEWCircuit TracerNEWSteerSAE EvalsExports Community BlogPrivacy & TermsContact
    1. Home
    3. Gemma-3-1B
    5. 22-GEMMASCOPE-2-RES-16K
    7. 10248
    Prev
    Next
    INDEX
    Explanations

    que se preguntan

    np_acts-logits-general · gemini-2.5-flash-lite

    text written in Romance languages (especially Italian, Spanish, Portuguese, and French).

    oai_token-act-pair · gpt-5Triggered by @smilgaveronika

    The neuron detects mentions of participant groups, clinical diagnoses, and study-related subject labels (e.g., ASD, NLD, ADHD, participants, group).

    oai_token-act-pair · gpt-5-miniTriggered by @smilgaveronika

    The neuron strongly activates on Italian nouns (content words like subjects, objects, and entities).

    oai_token-act-pair · o4-miniTriggered by @smilgaveronika

    references to ASD and neurodevelopmental disorders and discussions of global versus local visuospatial processing.

    oai_token-act-pair · gpt-5-nanoTriggered by @smilgaveronika
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    google/gemma-scope-2-1b-pt/resid_post/layer_22_width_16k_l0_medium
    Prompts (Dashboard)
    392,802 prompts, 256 tokens each
    Dataset (Dashboard)
    monology/pile-uncopyrighted
    No Configuration Found
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
     was
    2.03
     is
    1.70
     does
    1.67
     wasn
    1.63
     corresponds
    1.60
    was
    1.57
     occurs
    1.55
     reflects
    1.54
     isn
    1.50
     represents
    1.49
    POSITIVE LOGITS
     đều
    2.38
    都
    1.96
    都不
    1.92
    都很
    1.89
     vogliono
    1.87
     querem
    1.78
     possuem
    1.73
     ovat
    1.73
    都有
    1.73
     都
    1.72
    Activations Density 0.017%

    No Known Activations