© 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. 24-GEMMASCOPE-2-RES-16K
    4. 6352
    Prev
    Next
    INDEX
    Explanations

    The neuron fires on instructional or advisory language—phrases that address the reader with “you,” “make,” “choose,” “can,” etc., i.e. procedural advice.

    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_24_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
     
    0.29
     নারায়ণ
    0.29
     ±
    0.28
     irrespective
    0.27
     LPG
    0.27
    estream
    0.27
     ৭৭
    0.27
    µ
    0.27
    ographically
    0.27
    omechanical
    0.26
    POSITIVE LOGITS
     estadounidenses
    0.30
    让他们
    0.30
     problemy
    0.30
    䜣
    0.29
     Außerdem
    0.29
     मस्तिष्क
    0.29
     onları
    0.29
    ᒍ
    0.29
    ัญหา
    0.28
     problemler
    0.28
    Activations Density 0.051%

    No Known Activations