© 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. 34-GEMMASCOPE-RES-131K
    7. 32788
    Prev
    Next
    INDEX
    Explanations

    telephone routing and numbers

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

    The neuron fires most strongly on the words in the actual question titles or headings (e.g. “How can I map extensions to outbound routes” or “How to Find Cheap Calls to Bhutan…”), effectively detecting the question‐title lines.

    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
     phone
    -1.07
     telephone
    -0.98
     calls
    -0.82
    电话
    -0.82
     Phone
    -0.80
     call
    -0.80
    āci
    -0.77
    📞
    -0.76
     conversations
    -0.75
     dazzling
    -0.74
    POSITIVE LOGITS
    スカート
    0.84
     النها
    0.84
     tena
    0.81
     tand
    0.79
    itten
    0.77
    paccio
    0.77
     Jinping
    0.76
     ماش
    0.75
     Winchester
    0.75
     zama
    0.75
    Activations Density 0.020%

    No Known Activations