© 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. 7409
    Prev
    Next
    INDEX
    Explanations

    assumptions and reasoning

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

    The neuron isn’t picking out any coherent linguistic pattern but instead spikes on rare or “outlier” content words (i.e. infrequently seen tokens).

    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
     capitalization
    0.78
     eventual
    0.77
    rául
    0.77
    ি
    0.77
    ي
    0.74
    ра
    0.73
    ارب
    0.70
     inquisitive
    0.68
     anarchy
    0.68
     conservatively
    0.68
    POSITIVE LOGITS
     './
    0.77
     nombr
    0.75
     ⠀
    0.74
    🐧
    0.71
    🦑
    0.71
    leader
    0.71
    navigate
    0.70
    control
    0.70
    shine
    0.69
     『
    0.69
    Activations Density 0.000%

    No Known Activations