© 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. 15161
    Prev
    Next
    INDEX
    Explanations

    usage of them

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

    This neuron flags content‐bearing (non–stop) words—i.e. relatively rare or information-rich tokens rather than common function words.

    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
    ↵↵
    -2.13
    }$
    -2.11
     it
    -2.09
    柃
    -2.05
    fordshire
    -2.03
    '
    -1.98
     zijn
    -1.97
    ၟ
    -1.95
    Blaðsíða
    -1.95
     atau
    -1.95
    POSITIVE LOGITS
    苼
    3.20
     to
    2.83
     It
    2.81
    は
    2.34
    loses
    2.33
    geox
    2.20
     themselves
    2.17
    seemed
    2.17
    訫
    2.08
    wouldn
    2.06
    Activations Density 0.045%

    No Known Activations