© 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. 113759
    Prev
    Next
    INDEX
    Explanations

    input fields

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

    HTML input elements with type="text".

    oai_token-act-pair · claude-4-5-sonnetTriggered by @bfgcn7frf5

    The neuron fires on occurrences of the HTML input “text” type attribute—i.e. it spots input fields whose type is “text.”

    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
    蓠
    -2.27
     ส์
    -2.20
    轳
    -2.13
     hugely
    -2.11
     depois
    -2.09
     différents
    -2.08
     acidente
    -2.08
     ridiculously
    -2.00
    nearly
    -1.98
     revamped
    -1.98
    POSITIVE LOGITS
     with
    2.58
     through
    2.52
     on
    2.47
     while
    2.28
     like
    2.20
     for
    2.19
     or
    2.09
     from
    1.98
     without
    1.94
     this
    1.91
    Activations Density 0.006%

    No Known Activations