© 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. 24-GEMMASCOPE-2-RES-16K
    7. 3712
    Prev
    Next
    INDEX
    Explanations

    statement release PR

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

    This neuron detects press‐release dateline and date metadata lines (numbers, slashes, and related punctuation in header datelines).

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

    text from press releases or newswire articles, particularly structural elements like location headers, dates, and the "/" separator before "PRNewswire" or "BUSINESS WIRE".

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

    press-release/newswire boilerplate, especially datelines with all-caps locations, dates, slashes, and wire-service attributions.

    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_24_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
     ஆனால்
    1.02
    Although
    0.97
    kość
    0.97
    This
    0.91
    goài
    0.89
     যদিও
    0.86
    dj
    0.86
    ایہ
    0.86
    While
    0.84
    də
    0.84
    POSITIVE LOGITS
     PR
    0.83
     USO
    0.82
     opportun
    0.81
     announce
    0.80
     underserved
    0.80
     U
    0.78
     to
    0.77
     hon
    0.75
     unprecedented
    0.74
     announcing
    0.74
    Activations Density 0.002%

    No Known Activations