© Neuronpedia 2026
    Privacy & TermsBlogGitHubSlackTwitterContact
    Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    Natural Language
    Autoencoders
    NEW
    Assistant AxisNEWCircuit TracerUPDATESteerSAE EvalsExportsAPI Community BlogPrivacy & TermsContact
    1. Home
    2. Gemma-3-12B
    3. 31-GEMMASCOPE-2-RES-16K
    4. 185
    Prev
    Next
    INDEX
    Explanations

    it can

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

    The neuron fires strongly on occurrences of the pronoun “it,” especially when it begins a sentence.

    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
    と共に
    0.40
     đều
    0.35
     lainnya
    0.32
     ,
    0.31
    2
    0.30
     etc
    0.30
    0
    0.30
    といった
    0.29
     
    0.29
     which
    0.28
    POSITIVE LOGITS
     seems
    0.62
     happens
    0.61
     beho
    0.55
     isn
    0.54
     goes
    0.54
     occurs
    0.54
     dawned
    0.54
     involves
    0.53
     feels
    0.52
     wasn
    0.52
    Activations Density 0.874%

    No Known Activations