© Neuronpedia 2026
    Privacy & TermsBlogGitHubSlackTwitterContact
    Neuronpedia logo - a computer chip with a rounded viewfinder border around it
    Neuronpedia
    Jacobian LensNEW
    Natural Language
    Autoencoders
    MAY 26    Assistant AxisJAN 26Circuit TracerAPR 26SteerSAE EvalsExportsinterp-engineNEW Community BlogPrivacy & TermsContact
    1. Home
    3. Gemma-2-27B
    5. 34-GEMMASCOPE-RES-131K
    7. 47413
    Prev
    Next
    INDEX
    Explanations

    crimes and offences

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

    The main thing this neuron does is spot words and phrases that signal legal accusations, admissions of guilt, or criminal responsibility.

    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
    不由得
    -0.91
     fatores
    -0.88
     Edited
    -0.87
    ovej
    -0.86
     智能
    -0.86
     项目
    -0.85
     antidesliz
    -0.81
     照片
    -0.81
     syst
    -0.79
     Aspects
    -0.79
    POSITIVE LOGITS
     crime
    3.73
     crimes
    3.61
    crime
    2.88
     Crime
    2.41
    Crime
    2.36
    crimes
    2.31
     Crimes
    1.93
     CRIME
    1.91
     wrongdoing
    1.79
     delitos
    1.74
    Activations Density 0.038%

    No Known Activations