© 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
    2. Gemma-2-27B
    3. 10-GEMMASCOPE-RES-131K
    4. 21712
    Prev
    Next
    INDEX
    Explanations

    introducing explanations or reasons

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

    The neuron fires on informal discourse markers (e.g. “Well,” or “Yeah,”) used to kick off a response.

    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
    因为
    -1.48
    6
    -1.43
    anță
    -1.38
    multer
    -1.38
    4
    -1.36
    "]));
    -1.36
    ですか
    -1.35
    7
    -1.34
     الذي
    -1.31
    5
    -1.31
    POSITIVE LOGITS
     like
    1.29
    顔を
    1.26
     DÍA
    1.24
    鷽
    1.23
     by
    1.22
     between
    1.21
    ignore
    1.21
    ciaio
    1.20
    こういった
    1.20
    鰱
    1.20
    Activations Density 0.022%

    No Known Activations