© 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. 128405
    Prev
    Next
    INDEX
    Explanations

    The neuron flags file‐path or URL fragments (text containing “/”) that point to resources like images, favicons, etc.

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

    resource paths like images

    np_acts-logits-general · gemini-2.5-flash-lite
    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
     there
    -1.29
     for
    -1.27
     those
    -1.18
     September
    -1.11
     that
    -1.09
     zawsze
    -1.04
     October
    -1.03
    γρά
    -1.03
     December
    -1.00
     at
    -1.00
    POSITIVE LOGITS
     latine
    1.42
     durer
    1.33
     maniere
    1.30
    镲
    1.23
     patienter
    1.14
     perpé
    1.11
     två
    1.10
     بدون
    1.08
     sedi
    1.06
     türkiye
    1.05
    Activations Density 0.005%

    No Known Activations