© 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
    3. Gemma-2-27B
    5. 22-GEMMASCOPE-RES-131K
    7. 67015
    Prev
    Next
    INDEX
    Explanations

    The King of glory

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

    This neuron detects paragraph or section breaks (i.e. the boundaries between blocks of text) and signals the length/count token that follows each block.

    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
     toks
    -1.19
     jego
    -1.16
     then
    -1.11
    べし
    -1.09
    même
    -1.09
    UNRELATED
    -1.06
    乁
    -1.06
     illustrazione
    -1.06
    ⅓
    -1.04
    もあるので
    -1.03
    POSITIVE LOGITS
     the
    1.70
     wykonane
    1.24
    哢
    1.19
    玙
    1.16
     our
    1.16
     OGSÅ
    1.13
     gustaba
    1.12
    ኹ
    1.10
     sostiene
    1.09
    彼女は
    1.09
    Activations Density 0.036%

    No Known Activations