Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    APIAssistant AxisNEWCircuit TracerNEWSteerSAE EvalsExports Community BlogPrivacy & TermsContact
    © Neuronpedia 2025
    Privacy & TermsBlogGitHubSlackTwitterContact
    1. Home
    2. Gemma-3-12B
    3. 12-GEMMASCOPE-2-RES-16K
    4. 196
    Prev
    Next
    INDEX
    Explanations

    is followed by a, always, huge

    np_acts-logits-general · gemini-2.5-flash-lite
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    google/gemma-scope-2-12b-pt/resid_post/layer_12_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
    س
    1.64
    p
    1.54
    Р
    1.39
    o
    1.34
    ت
    1.26
    Yea
    1.20
    т
    1.20
    an
    1.19
    же
    1.18
    m
    1.15
    POSITIVE LOGITS
     tincture
    1.02
     ’
    0.97
     pamoja
    0.96
     reign
    0.96
    entropic
    0.95
     tantamount
    0.95
    بار
    0.91
     twofold
    0.91
    opter
    0.91
    하다
    0.91
    Activations Density 1.178%

    No Known Activations