© 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. 26198
    Prev
    Next
    INDEX
    Explanations

    drug, gaga, download, carbon

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

    This neuron fires on uncommon, content‐bearing tokens (e.g. rare proper names or domain‐specific words) rather than on high-frequency 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
     finca
    -0.89
     сериал
    -0.84
    Dental
    -0.81
     biograf
    -0.79
    tives
    -0.78
    净化
    -0.76
     pido
    -0.76
     fré
    -0.76
    tismo
    -0.76
    Processing
    -0.75
    POSITIVE LOGITS
     cumul
    0.84
     oof
    0.81
     ürünleri
    0.79
    んび
    0.78
     ansch
    0.78
    meni
    0.77
    herren
    0.76
    臉
    0.75
    getS
    0.75
     marginBottom
    0.75
    Activations Density 0.030%

    No Known Activations