© 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. 65488
    Prev
    Next
    INDEX
    Explanations

    question openers and technical roles

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

    The neuron flags long numeric (often hexadecimal) sequences in code or comments.

    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
     datter
    -1.16
     hadas
    -1.11
     kapital
    -1.09
     farine
    -1.05
     konflikt
    -1.02
     décide
    -1.02
     dépens
    -1.02
     réfrig
    -1.02
     pittores
    -1.02
    ilización
    -1.01
    POSITIVE LOGITS
     brimming
    0.90
    om
    0.90
     now
    0.85
     advierte
    0.85
     John
    0.84
    am
    0.83
     Was
    0.82
     consigue
    0.81
     I
    0.80
     Man
    0.79
    Activations Density 0.001%

    No Known Activations