© 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-3-12B
    5. 31-GEMMASCOPE-2-RES-16K
    7. 13162
    Prev
    Next
    INDEX
    Explanations

    parallel lines have so much in common

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

    The neuron strongly activates on unusual or non-ASCII unicode character tokens (e.g., accented letters or foreign script).

    oai_token-act-pair · o4-miniTriggered by @jyhe0408
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    google/gemma-scope-2-12b-pt/resid_post/layer_31_width_16k_l0_medium
    Prompts (Dashboard)
    392,802 prompts, 256 tokens each
    Dataset (Dashboard)
    monology/pile-uncopyrighted
    No Configuration Found
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
    клар
    0.56
    クライ
    0.54
     配
    0.53
    覬
    0.52
    贊
    0.52
    leicht
    0.51
     Zub
    0.51
    വിൽ
    0.50
     Dign
    0.50
    currentChar
    0.50
    POSITIVE LOGITS
    uthers
    0.59
    ##
    0.59
     ὑ
    0.57
     IBS
    0.56
    𝑦
    0.56
    Click
    0.56
    𝐡
    0.55
     OSX
    0.54
    ɾ
    0.54
    When
    0.54
    Activations Density 0.016%

    No Known Activations