© 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. 107667
    Prev
    Next
    INDEX
    Explanations

    drugs and illicit substances

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

    vocabulary related to illicit drugs and substance use treatment, particularly opioids, cannabis, and drug-related legal issues.

    oai_token-act-pair · claude-3-7-sonnet-20250219Triggered by @neilrathi

    The neuron activates on numerical tokens such as digits, years, and other numbers in the text.

    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
     procesador
    -0.78
    lâ
    -0.71
    TAS
    -0.71
    Discipline
    -0.71
    ctos
    -0.71
     asc
    -0.70
     SATA
    -0.70
     lâ
    -0.69
     federal
    -0.69
    +-
    -0.69
    POSITIVE LOGITS
     harms
    1.10
     cannabis
    0.94
     injecting
    0.87
     Inject
    0.86
     dependence
    0.81
     sharps
    0.80
    harm
    0.80
     harm
    0.79
     psychosis
    0.78
     TikTok
    0.77
    Activations Density 0.068%

    No Known Activations