© 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. 34-GEMMASCOPE-RES-131K
    7. 63224
    Prev
    Next
    INDEX
    Explanations

    Html form method post

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

    This neuron fires on programming‐style tokens—especially PascalCase identifiers, API/function names, and other code symbols.

    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
     from
    -0.95
     even
    -0.94
     including
    -0.91
     those
    -0.88
     on
    -0.88
     more
    -0.85
     outweigh
    -0.85
     maybe
    -0.83
    流通
    -0.82
    なります
    -0.82
    POSITIVE LOGITS
     трени
    0.93
     Amerikaanse
    0.90
     gib
    0.90
    pflicht
    0.90
    錮
    0.87
     raconter
    0.85
     FLASH
    0.84
     verschillende
    0.82
     salts
    0.82
     Elevator
    0.81
    Activations Density 0.001%

    No Known Activations