© 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. 31-GEMMASCOPE-2-RES-16K
    7. 2613
    Prev
    Next
    INDEX
    Explanations

    explain roleplay scenarios

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

    This neuron fires strongly at the beginnings (and throughout) of distinct document sections or “blocks,” effectively marking and measuring section boundaries.

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

    paragraph or section breaks in structured informational or instructional text.

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

    large multi-digit decimal numbers (floating-point values) embedded in the text.

    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_31_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
     The
    1.08
    The
    1.03
    C
    1.02
    A
    1.00
     It
    0.97
    B
    0.96
    F
    0.96
     If
    0.91
    It
    0.91
    This
    0.91
    POSITIVE LOGITS
    ↵↵↵↵↵↵↵↵↵↵↵↵↵↵↵↵↵↵↵↵↵↵↵↵↵↵↵↵↵
    0.82
    ↵↵↵↵↵↵↵↵↵↵↵
    0.81
     excellently
    0.80
    ↵↵↵↵↵↵↵↵↵↵↵↵↵↵↵↵↵↵↵↵↵↵↵
    0.80
    ↵↵↵↵↵↵↵
    0.78
    ↵↵↵↵↵↵↵↵↵↵↵↵↵↵↵↵↵↵↵↵↵↵↵↵↵↵↵
    0.78
    <start_of_image>
    0.77
    ↵↵↵↵↵↵↵↵↵↵↵↵↵↵↵
    0.75
    ↵↵↵↵↵↵↵↵↵↵↵↵↵
    0.74
    ’.”
    0.74
    Activations Density 0.067%

    No Known Activations