© 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. 1155
    Prev
    Next
    INDEX
    Explanations

    The neuron is tuned to pick up formal‐discourse tokens—especially institutional or abstract terms like “state” (and associated high-register pronouns/auxiliaries such as “this” and “has”) in expository passages.

    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
    olds
    0.78
    uie
    0.78
    ایل
    0.77
    Gestion
    0.77
    ہنی
    0.76
    Rpc
    0.75
    胄
    0.75
    CodeDict
    0.75
    orpion
    0.74
    臨時
    0.74
    POSITIVE LOGITS
     pector
    0.77
    ১
    0.77
    ㅏ
    0.75
     steg
    0.73
     unlike
    0.73
     Unlike
    0.73
     album
    0.72
    으로써
    0.72
     mortgagee
    0.72
    			
    0.72
    Activations Density 0.000%

    No Known Activations