© Neuronpedia 2026
    Privacy & TermsBlogGitHubSlackTwitterContact
    Neuronpedia logo - a computer chip with a rounded viewfinder border around it
    Neuronpedia
    Jacobian LensNEW
    Natural Language
    Autoencoders
    MAY 26    Assistant AxisJAN 26Circuit TracerAPR 26SteerSAE EvalsExportsinterp-engineNEW Community BlogPrivacy & TermsContact
    1. Home
    3. Gemma-2-27B
    5. 22-GEMMASCOPE-RES-131K
    7. 128013
    Prev
    Next
    INDEX
    Explanations

    The neuron strongly activates whenever it sees the word “things.”

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

    things followed by description

    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
    вгений
    -2.25
    荦
    -2.23
     撮っ
    -2.22
    鈊
    -2.22
    蟊
    -2.22
     oorsp
    -2.19
     komende
    -2.19
    to
    -2.19
    一個
    -2.17
    簓
    -2.16
    POSITIVE LOGITS
    </h3>
    2.84
     могут
    2.36
     to
    2.31
     (
    2.30
     „
    2.30
    </em>
    2.28
    </strong>
    2.23
    {
    2.17
     in
    2.14
    .
    2.13
    Activations Density 0.024%

    No Known Activations