© 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. 4209
    Prev
    Next
    INDEX
    Explanations

    special and general relativity

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

    The neuron strongly fires on key formal/legal or sales‐type “action” words—especially verbs and nouns like contract, concluded, provides, offers, between—that signal obligations, transfers, or amenity offerings in a formal or listing context.

    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_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
    E
    0.89
    O
    0.86
    ur
    0.84
    sky
    0.79
    s
    0.78
    urie
    0.74
    sl
    0.74
    K
    0.74
    A
    0.73
    urbo
    0.73
    POSITIVE LOGITS
     audioElement
    0.84
     mildly
    0.82
     moderately
    0.78
     houd
    0.77
    命令行
    0.77
     Navbar
    0.77
    インテリア
    0.76
     humbly
    0.74
    旂
    0.74
     ziff
    0.73
    Activations Density 0.001%

    No Known Activations