© 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. 117171
    Prev
    Next
    INDEX
    Explanations

    hypocrisy

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

    The main thing this neuron does is pick out pejorative or insulting descriptors (i.e. words used in personal attacks).

    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
     zwar
    -0.88
    otheek
    -0.85
     optimistic
    -0.84
    еро
    -0.84
     supernatural
    -0.82
     mistakes
    -0.82
    俚
    -0.81
    tión
    -0.79
     troubleshooting
    -0.79
    rations
    -0.78
    POSITIVE LOGITS
     hypocrisy
    4.38
    Hypo
    4.00
     hypocritical
    3.98
     Hypo
    3.91
    hypo
    3.89
     hypo
    3.78
     hypocrite
    3.67
     hipo
    2.41
     inconsistency
    2.13
     hipó
    1.91
    Activations Density 0.040%

    No Known Activations