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

    Neuronpedia

    Jacobian LensNEW
    Natural Language
    Autoencoders
    NEW    Assistant AxisNEWCircuit TracerUPDATESteerSAE EvalsExportsinterp-engineNEWAPI Community BlogPrivacy & TermsContact
    1. Home
    3. Gemma-3-12B
    5. 24-GEMMASCOPE-2-RES-16K
    7. 3461
    Prev
    Next
    INDEX
    Explanations

    resource provides

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

    This neuron strongly responds to product‐specification and marketing terms—i.e. nouns and adjectives listing item attributes (colors, styles, materials, quantities) in a commercial/product‐description context.

    oai_token-act-pair · o4-miniTriggered by @jyhe0408

    commas in product descriptions and promotional material.

    oai_token-act-pair · claude-4-5-sonnetTriggered by @jyhe0408

    promotional or product-description language that highlights features, materials, qualities, and specifications.

    oai_token-act-pair · gpt-5Triggered by @jyhe0408
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    google/gemma-scope-2-12b-pt/resid_post/layer_24_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
     그러면
    0.73
     لأنه
    0.70
     因為
    0.70
     لأن
    0.64
     경우에는
    0.64
     omdat
    0.64
    그러
    0.64
     পরিপ্রেক্ষিতে
    0.63
     因为
    0.63
     그러면은
    0.60
    POSITIVE LOGITS
     biedt
    0.84
     awaits
    0.79
    始终
    0.74
     bietet
    0.71
    一直是
    0.71
    如今
    0.70
     ofrece
    0.65
     proclaims
    0.65
     mutta
    0.64
     comes
    0.64
    Activations Density 0.679%

    No Known Activations