© 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. 10-GEMMASCOPE-RES-131K
    7. 103349
    Prev
    Next
    INDEX
    Explanations

    saying, telling, hearing

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

    This neuron detects words that signal or report acts of speaking or writing (speech‐act verbs and quotation markers).

    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
    峮
    -1.36
     Hierro
    -1.32
    verhält
    -1.32
     estrada
    -1.31
     rouges
    -1.29
    拮
    -1.27
     vitesse
    -1.23
     sauter
    -1.23
    ён
    -1.20
    дови
    -1.20
    POSITIVE LOGITS
     about
    3.11
     before
    1.73
     they
    1.55
     that
    1.46
     Informações
    1.42
     będ
    1.37
     when
    1.34
     each
    1.34
    5
    1.31
     właścic
    1.30
    Activations Density 0.030%

    No Known Activations