© 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. 1150
    Prev
    Next
    INDEX
    Explanations

    Australian places and history

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

    The neuron strongly activates on newspaper‐style datelines and location headings (e.g. uppercase place names and similar line headers).

    oai_token-act-pair · o4-miniTriggered by @jyhe0408

    optical character recognition (OCR) errors and garbled text.

    oai_token-act-pair · claude-4-5-sonnetTriggered by @jyhe0408

    capitalized named entities in formal records, such as people’s names with honorifics/initials, place and institution names, dates, and titles typical of obituaries, biographies, and bibliographies.

    oai_token-act-pair · gpt-5Triggered 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
     использу
    0.94
     поскольку
    0.91
     planar
    0.88
     charismatic
    0.88
     RMSE
    0.88
     genauso
    0.86
     imprecise
    0.86
     mindset
    0.83
     NOAA
    0.83
     badass
    0.83
    POSITIVE LOGITS
     Messrs
    1.11
    ,—
    1.06
    ;,
    0.95
    Í
    0.95
    Railway
    0.95
    Queensland
    0.88
    fc
    0.85
     Railway
    0.85
    ,-
    0.83
    .—
    0.83
    Activations Density 0.091%

    No Known Activations