© 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. 24-GEMMASCOPE-2-RES-16K
    7. 15342
    Prev
    Next
    INDEX
    Explanations

    sewing machine, Chat Commands, The Machine

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

    The neuron spikes on clause-initial subjects and auxiliaries (e.g. “it,” “they,” “could”)—in other words, the opening word of a new main clause.

    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
     terminates
    0.84
     counselor
    0.81
     pediatrician
    0.79
     Counselor
    0.77
     envisions
    0.75
    л
    0.74
     administers
    0.74
     bookstore
    0.73
    ни
    0.73
     certifies
    0.72
    POSITIVE LOGITS
     Questi
    0.77
    Produtos
    0.76
    rzym
    0.74
     posibil
    0.74
     쉽
    0.73
    普遍
    0.72
     Ciò
    0.71
     compart
    0.71
     సా
    0.70
    SendData
    0.70
    Activations Density 0.000%

    No Known Activations