© 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. 129133
    Prev
    Next
    INDEX
    Explanations

    This neuron fires on the very first token of each new text block (i.e. it marks the start‐of‐block or first‐word position).

    oai_token-act-pair · o4-miniTriggered by @jyhe0408

    getChildAt

    np_acts-logits-general · gemini-2.5-flash-lite
    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
     of
    -1.13
     thier
    -0.97
    岡田
    -0.87
    през
    -0.87
     eikä
    -0.87
     kanilang
    -0.82
    𝟷
    -0.82
     می‌باشد
    -0.81
     ihre
    -0.81
     that
    -0.80
    POSITIVE LOGITS
    팹
    1.12
    ktora
    1.05
     MonoBehaviour
    1.00
     Depois
    0.99
    parseDouble
    0.97
    vVar
    0.96
    ƚ
    0.96
     penultimate
    0.95
     Lleva
    0.94
    Facades
    0.94
    Activations Density 0.221%

    No Known Activations