© 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. 31-GEMMASCOPE-2-RES-16K
    7. 12434
    Prev
    Next
    INDEX
    Explanations

    narrow or wide

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

    The neuron is detecting words at the very start of new sections or sentences (e.g. the first token in a paragraph).

    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
     선생님
    0.67
    गो
    0.66
    pym
    0.65
    Zn
    0.63
    𝜁
    0.61
     coût
    0.60
     Certains
    0.59
    골
    0.59
    ब्बा
    0.59
     Ärz
    0.59
    POSITIVE LOGITS
     extremes
    0.73
     wide
    0.72
     broad
    0.63
     ranges
    0.61
     range
    0.60
     abrang
    0.59
    幅広い
    0.58
     fasc
    0.58
     extreme
    0.57
     breadth
    0.57
    Activations Density 0.001%

    No Known Activations