© 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 EvalsExportsAPI Community BlogPrivacy & TermsContact
    1. Home
    3. Gemma-2-27B
    5. 34-GEMMASCOPE-RES-131K
    7. 49155
    Prev
    Next
    INDEX
    Explanations

    Made my

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

    This neuron is selectively responding to the leading whitespace/indentation tokens in source code, i.e. it detects indent depth in code.

    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
     suivie
    -0.73
     kó
    -0.71
     Plan
    -0.69
     Challenges
    -0.69
    ANIE
    -0.68
     hadn
    -0.68
     именно
    -0.67
    Joint
    -0.67
     Mag
    -0.66
     września
    -0.65
    POSITIVE LOGITS
    agi
    0.82
     наступ
    0.80
     Payroll
    0.77
    odorant
    0.77
    acij
    0.75
    ต่อไป
    0.75
    Ense
    0.75
    jamos
    0.74
    Eventos
    0.74
    assination
    0.74
    Activations Density 0.005%

    No Known Activations