© 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
    2. Gemma-2-27B
    3. 34-GEMMASCOPE-RES-131K
    4. 93078
    Prev
    Next
    INDEX
    Explanations

    drug packaging, treatment, design, study, synthesis, manufacture

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

    phrases starting with "for the" in scientific or medical contexts.

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

    The neuron is sensitive to very high‐frequency tokens—common function words and other extremely frequent terms.

    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
    身子
    -1.07
    つける
    -1.05
     белом
    -1.00
     prawa
    -1.00
    努力
    -1.00
     artesanía
    -1.00
     consolidating
    -1.00
    骉
    -1.00
     trzech
    -0.98
     informace
    -0.97
    POSITIVE LOGITS
     herbes
    1.09
     geforce
    1.05
     identification
    1.05
     detection
    1.05
     establishment
    1.05
    vää
    1.05
     vano
    1.04
     noten
    1.02
     implementation
    1.00
    Thankfully
    1.00
    Activations Density 0.278%

    No Known Activations