© 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. 10-GEMMASCOPE-RES-131K
    7. 14913
    Prev
    Next
    INDEX
    Explanations

    pronoun followed by verb/adverb

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

    This neuron fires on words appearing at the starts of sentences (i.e. sentence-initial tokens).

    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
    んだから
    -1.57
     opět
    -1.53
     permettent
    -1.53
    ütter
    -1.52
     zároveň
    -1.49
    máy
    -1.48
     engelsk
    -1.46
    новь
    -1.45
     this
    -1.45
     hendes
    -1.45
    POSITIVE LOGITS
     first
    1.42
     später
    1.34
     späteren
    1.26
     with
    1.26
    훗
    1.17
    Then
    1.16
     Initially
    1.16
    ׇ
    1.14
     ingredi
    1.13
     intended
    1.11
    Activations Density 0.138%

    No Known Activations