© 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. 65749
    Prev
    Next
    INDEX
    Explanations

    Posts Tagged, Tag Archives, Sawmill

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

    The neuron fires strongly on numeric tokens (especially multi‐digit figures, e.g. years or large counts).

    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
     Kartoffeln
    -0.98
    ución
    -0.98
    Démographie
    -0.93
    )]);
    -0.91
     färg
    -0.91
     naudo
    -0.91
    lanzen
    -0.90
    ânsito
    -0.89
     específicas
    -0.87
     täv
    -0.87
    POSITIVE LOGITS
     thier
    0.89
     a
    0.89
     различных
    0.85
     frü
    0.85
    onian
    0.82
    ische
    0.82
    穆
    0.79
    qai
    0.78
     seuls
    0.76
     некото
    0.76
    Activations Density 0.000%

    No Known Activations