© 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 EvalsExportsAPI Community BlogPrivacy & TermsContact
    1. Home
    3. Gemma-3-12B
    5. 31-GEMMASCOPE-2-RES-16K
    7. 12006
    Prev
    Next
    INDEX
    Explanations

    data structures

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

    the presence of relatively rare, content‐heavy tokens (i.e. uncommon nouns and verbs with high semantic weight)

    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
     as
    1.99
    作为
    1.66
     как
    1.64
     sebagai
    1.59
    作為
    1.59
     як
    1.52
     jako
    1.51
     作为
    1.47
     ως
    1.42
    作为一个
    1.42
    POSITIVE LOGITS
    并在
    0.81
     መጠን
    0.70
    异步
    0.62
     ஆனால்
    0.61
    并通过
    0.61
     并
    0.60
    并将
    0.60
    并
    0.58
     despite
    0.58
    便利
    0.58
    Activations Density 0.212%

    No Known Activations