© 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. 34-GEMMASCOPE-RES-131K
    7. 48610
    Prev
    Next
    INDEX
    Explanations

    API parameters and data

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

    The neuron responds strongly to digit‐containing tokens (numeric literals, IDs, version numbers, file sizes, etc.).

    oai_token-act-pair · o4-miniTriggered by @jyhe0408
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    google/gemma-scope-27b-pt-res/layer_34/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
     both
    -0.98
     those
    -0.95
    phalt
    -0.85
     still
    -0.84
    outils
    -0.83
    氾
    -0.80
     divertir
    -0.80
     would
    -0.79
    '");
    -0.78
    terly
    -0.77
    POSITIVE LOGITS
    cmd
    1.00
    method
    0.94
    ulaire
    0.88
    szön
    0.88
    token
    0.86
     method
    0.85
    Cmd
    0.84
    Authorization
    0.84
     είναι
    0.84
    JSON
    0.82
    Activations Density 0.020%

    No Known Activations