© Neuronpedia 2026
    Privacy & TermsBlogGitHubSlackTwitterContact
    Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    Natural Language
    Autoencoders
    NEW    Assistant AxisNEWCircuit TracerUPDATESteerSAE EvalsExportsAPI Community BlogPrivacy & TermsContact
    1. Home
    3. Gemma-3-12B
    5. 31-GEMMASCOPE-2-RES-16K
    7. 8312
    Prev
    Next
    INDEX
    Explanations

    `a href` or `"` onclick

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

    The neuron is specialized to spot HTML anchor elements and raw URL strings (i.e., hyperlinks).

    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
     possibilités
    0.64
    '/>
    0.63
    ensione
    0.61
     />
    0.59
     ľ
    0.59
     boissons
    0.59
    ینے
    0.58
     capacitive
    0.58
    )}/>
    0.58
     luctus
    0.58
    POSITIVE LOGITS
    ))(
    0.72
    Rel
    0.68
    Target
    0.59
    ">{{
    0.58
    REL
    0.57
    ">${
    0.57
    Strong
    0.56
    Herm
    0.56
    >(_
    0.56
    Vari
    0.55
    Activations Density 0.005%

    No Known Activations