© 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
    2. Gemma-3-12B
    3. 31-GEMMASCOPE-2-RES-16K
    4. 6748
    Prev
    Next
    INDEX
    Explanations

    OR followed by parentheses

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

    The neuron isn’t responding to any particular word meaning or part of speech but rather to the transition into or presence of user‐generated “comment” style text (forum/blog comment language, punctuation and formatting) as opposed to the preceding formal/reporting style.

    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
     (')
    0.72
     []
    0.64
     {}
    0.62
    ('.')
    0.62
     ')')
    0.61
     '.')
    0.60
     ('
    0.59
    >>()
    0.59
     ()
    0.58
     [][]
    0.58
    POSITIVE LOGITS
    ೆಯನ್ನು
    0.76
    utom
    0.68
     بہترین
    0.64
    abet
    0.62
    জনকে
    0.62
    жин
    0.61
    ຫນ
    0.61
    金を
    0.61
    ವನ್ನು
    0.61
    チャレンジ
    0.61
    Activations Density 0.205%

    No Known Activations