© 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. 22-GEMMASCOPE-RES-131K
    7. 15469
    Prev
    Next
    INDEX
    Explanations

    list indices and positions

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

    The neuron fires on integer-related tokens (e.g. numeric literals and the keyword “int”) in code.

    oai_token-act-pair · o4-miniTriggered by @jyhe0408

    Programming-related terms for position, index, or row identifiers in code.

    oai_token-act-pair · claude-4-5-sonnetTriggered by @jyhe0408

    references to row or item indices/positions in lists or UI components.

    oai_token-act-pair · gpt-5Triggered by @jyhe0408
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    google/gemma-scope-27b-pt-res/layer_22/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
     Souver
    -1.00
     and
    -0.99
    ड़ी
    -0.98
     diyor
    -0.94
     about
    -0.93
     as
    -0.89
    :
    -0.89
     each
    -0.88
     \
    -0.88
    DETAILED
    -0.87
    POSITIVE LOGITS
    complished
    1.16
    Kenmerken
    1.15
    瞅
    1.13
    ditionally
    1.06
     RemoteException
    1.04
     Witkin
    1.03
    笵
    1.03
    minated
    1.03
    wię
    1.01
     бампер
    1.00
    Activations Density 0.007%

    No Known Activations