© 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. 22-GEMMASCOPE-RES-131K
    4. 34353
    Prev
    Next
    INDEX
    Explanations

    programming context

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

    The neuron is looking for the kind of formal, “API‐style” documentation language that describes what a function or procedure does (e.g. “takes two arrays and returns the sum of their products,” “attaches a listener,” “I’ll first get an array of arguments,” etc.).

    oai_token-act-pair · o4-miniTriggered 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
     poliéster
    -1.19
    Welke
    -1.09
     camisas
    -1.06
     kulinar
    -1.02
    
    -0.99
     adicion
    -0.97
    ๘
    -0.97
     ваз
    -0.96
     pomaga
    -0.96
     molho
    -0.96
    POSITIVE LOGITS
     francisco
    1.10
    😕
    1.05
     will
    1.01
     ♡
    0.99
     »>
    0.95
    😚
    0.95
    when
    0.95
    屏蔽
    0.94
     from
    0.94
     sebagainya
    0.93
    Activations Density 0.045%

    No Known Activations