© 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. 12620
    Prev
    Next
    INDEX
    Explanations

    shielding, filtering, forming

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

    The neuron strongly fires on second-person, reader-addressing terms and constructions (e.g. “you,” “your,” and related context).

    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
    0.52
     Crypto
    0.42
     dubbed
    0.42
     C
    0.40
     +
    0.39
     reportedly
    0.39
    7
    0.39
    ﾞ
    0.39
     seven
    0.36
     Dec
    0.36
    POSITIVE LOGITS
    thisobject
    0.45
    readdir
    0.44
     dresser
    0.43
     положение
    0.43
    秏
    0.43
     mirar
    0.43
     понятие
    0.43
    parvec
    0.43
    왠
    0.43
     equipments
    0.42
    Activations Density 0.377%

    No Known Activations