© 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. 98948
    Prev
    Next
    INDEX
    Explanations

    mediational role

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

    The neuron is detecting URL path tokens—slashes and the directory‐ or resource names (like “artist”) in web addresses.

    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
    Dalam
    -0.91
    ܥ
    -0.90
     Traduction
    -0.84
     générer
    -0.84
    ectores
    -0.82
     VICTORIA
    -0.82
    uco
    -0.81
     LYON
    -0.81
    িয়ে
    -0.81
     العرب
    -0.81
    POSITIVE LOGITS
     МВД
    1.09
    厚的
    1.05
    SearchResult
    0.91
    pgterms
    0.90
     biographical
    0.89
     remain
    0.88
     people
    0.88
     searched
    0.88
     search
    0.88
     names
    0.88
    Activations Density 0.005%

    No Known Activations