© 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
    2. Gemma-2-27B
    3. 22-GEMMASCOPE-RES-131K
    4. 129545
    Prev
    Next
    INDEX
    Explanations

    This neuron fires on short tokens that look like technical identifiers or numeric codes—e.g. chunks of hex strings, version numbers, citation numerals, or similar digit‐heavy/ID‐style tokens.

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

    filenames and code delimiters

    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
     for
    -1.66
     This
    -1.48
     For
    -1.41
     or
    -1.34
     their
    -1.29
    為に
    -1.26
    ようです
    -1.22
     If
    -1.21
     the
    -1.20
     on
    -1.20
    POSITIVE LOGITS
    がありますが
    1.49
    もあります
    1.41
    なども
    1.38
    もあって
    1.35
    があります
    1.35
    などは
    1.33
    もありました
    1.33
    も多く
    1.27
     volant
    1.21
    もありますが
    1.21
    Activations Density 0.001%

    No Known Activations