© Neuronpedia 2026
    Privacy & TermsBlogGitHubSlackTwitterContact
    Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    Natural Language
    Autoencoders
    NEW    Assistant AxisNEWCircuit TracerUPDATESteerSAE EvalsExportsAPI Community BlogPrivacy & TermsContact
    1. Home
    3. Gemma-2-27B
    5. 10-GEMMASCOPE-RES-131K
    7. 70672
    Prev
    Next
    INDEX
    Explanations

    direction and facing

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

    The neuron activates on words that denote spatial orientation or direction (e.g. downwards, downward, north).

    oai_token-act-pair · o4-miniTriggered by @jyhe0408
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    google/gemma-scope-27b-pt-res/layer_10/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
     into
    -1.65
    嘖
    -1.48
    小声
    -1.48
    瞜
    -1.47
    スポンサーリンク
    -1.47
     joked
    -1.46
     famoso
    -1.45
     the
    -1.45
     surgir
    -1.45
     spezielle
    -1.43
    POSITIVE LOGITS
     if
    1.63
    ^{
    1.45
    
    1.45
     \"
    1.36
     ""
    1.34
    mıştır
    1.34
    erráneo
    1.34
     вашего
    1.33
    。
    1.32
    indest
    1.31
    Activations Density 0.009%

    No Known Activations