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

    Neuronpedia

    Natural Language
    Autoencoders
    NEW    Assistant AxisNEWCircuit TracerUPDATESteerSAE EvalsExportsAPI Community BlogPrivacy & TermsContact
    1. Home
    3. Gemma-2-27B
    5. 22-GEMMASCOPE-RES-131K
    7. 95765
    Prev
    Next
    INDEX
    Explanations

    sexual assault and misconduct

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

    the neuron identifies explicit references to sexual assault or abuse.

    oai_token-act-pair · o4-miniTriggered by @jyhe0408
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    google/gemma-scope-27b-pt-res/layer_22/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
     like
    -1.72
     now
    -1.28
     such
    -1.18
     of
    -1.14
    H
    -1.13
     will
    -1.09
     miliardi
    -1.09
     how
    -1.07
    !
    -1.05
    X
    -1.04
    POSITIVE LOGITS
     plusieurs
    1.55
    wanna
    1.44
    banez
    1.42
     déplacements
    1.39
    僰
    1.34
     actrices
    1.33
    aned
    1.30
    keda
    1.28
     regalos
    1.26
    odine
    1.25
    Activations Density 0.020%

    No Known Activations