© 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-3-27B-IT
    5. 3-GEMMASCOPE-2-TRANSCODER-262K
    7. 123165
    Prev
    Next
    INDEX
    Explanations

    The neuron seems to detect phrases indicating a state of activity, presence or movement, often followed by common sentence connectors.Here's a breakdown of my thought process:1. **being active or present**

    np_acts-logits-general · gemini-2.5-flash-lite
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    google/gemma-scope-2-27b-it/transcoder_all/layer_3_width_262k_l0_small_affine
    Prompts (Dashboard)
    238,145 prompts, 512 tokens each
    Dataset (Dashboard)
    lmsys + oasst1
    No Configuration Found
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
    ap
    0.90
    ur
    0.87
    were
    0.85
    AL
    0.81
    p
    0.80
    MA
    0.79
    ai
    0.76
    water
    0.76
    ED
    0.71
    ov
    0.70
    POSITIVE LOGITS
    יות
    0.91
    да
    0.79
    ات
    0.75
    ↵
    0.72
    '
    0.71
    ї
    0.71
    敀
    0.70
    ある
    0.70
    人
    0.69
    し
    0.69
    Activations Density 0.026%

    No Known Activations