© 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-2-27B
    5. 34-GEMMASCOPE-RES-131K
    7. 47460
    Prev
    Next
    INDEX
    Explanations

    hell or ell variants

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

    The neuron is triggered by occurrences of the substring “Hell” (e.g. in “Hell,” “Hell’s,” “Hellinger,” etc.).

    oai_token-act-pair · o4-miniTriggered by @jyhe0408
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    google/gemma-scope-27b-pt-res/layer_34/width_131k
    Prompts (Dashboard)
    24,576 prompts, 128 tokens each
    Dataset (Dashboard)
    monology/pile-uncopyrighted
    No Configuration Found
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
     traff
    -0.73
     посвящен
    -0.72
    ecte
    -0.71
    coop
    -0.70
     Kazan
    -0.69
     Wooden
    -0.69
     conspic
    -0.68
    wissen
    -0.67
    gear
    -0.66
    しかない
    -0.66
    POSITIVE LOGITS
    reté
    0.78
    Portale
    0.76
     bent
    0.74
    inger
    0.74
     Feind
    0.73
     feine
    0.72
     familien
    0.72
     demonios
    0.71
     klare
    0.71
    Dzi
    0.70
    Activations Density 0.035%

    No Known Activations