© 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. 24-GEMMASCOPE-2-RES-16K
    7. 13148
    Prev
    Next
    INDEX
    Explanations

    repeated broken promises

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

    The neuron activates strongly on semantically rich content words (nouns and verbs) rather than on common function words.

    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
    ,
    0.70
    .
    0.68
     detto
    0.67
     grammatical
    0.64
     appetizer
    0.64
     technical
    0.64
     صلی
    0.64
     general
    0.63
     grammat
    0.63
     et
    0.63
    POSITIVE LOGITS
    searchValue
    0.70
    istically
    0.68
    팠
    0.67
    езда
    0.64
    옅
    0.64
    <unused45>
    0.63
    ម្រ
    0.63
    IZED
    0.63
    <unused71>
    0.63
     выбирать
    0.63
    Activations Density 0.000%

    No Known Activations