© 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. 34-GEMMASCOPE-RES-131K
    7. 53845
    Prev
    Next
    INDEX
    Explanations

    breach, fraud, shooting, invitation

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

    The neuron strongly activates on language describing legal judgments and criminal charges (e.g., pleas, convictions, offences, sentences).

    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
    Caching
    -0.75
     KILL
    -0.75
     Canton
    -0.74
     SCN
    -0.74
    enough
    -0.72
     CONV
    -0.72
    Filing
    -0.72
     rezept
    -0.72
     احتمال
    -0.72
    zom
    -0.71
    POSITIVE LOGITS
    akat
    0.80
     armée
    0.76
    bes
    0.74
     breach
    0.74
     breaches
    0.73
     harian
    0.73
     liệu
    0.71
     Rel
    0.70
     Breach
    0.70
     couleurs
    0.69
    Activations Density 0.010%

    No Known Activations