© 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. 1700
    Prev
    Next
    INDEX
    Explanations

    work well

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

    The neuron strongly activates on the last content word before a sentence boundary.

    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
     embarazo
    0.56
     Jacq
    0.56
     virtud
    0.54
     turista
    0.54
     Jacqueline
    0.53
     tava
    0.53
     Pomer
    0.53
    딧
    0.53
     pornography
    0.52
     consecuencia
    0.52
    POSITIVE LOGITS
     通过
    0.70
     through
    0.68
    Level
    0.67
     niezbęd
    0.66
     melalui
    0.64
    通过
    0.64
    Includes
    0.62
    ustain
    0.60
     setiap
    0.60
     através
    0.59
    Activations Density 0.040%

    No Known Activations