© 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. 31-GEMMASCOPE-2-RES-16K
    7. 3152
    Prev
    Next
    INDEX
    Explanations

    The neuron flags unusually rare or low‐frequency tokens—i.e. standout, high–information‐content words.

    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
     induc
    0.51
    하다고
    0.50
    叔
    0.50
     अधि
    0.50
    ())),
    0.50
     covariate
    0.50
     נק
    0.48
     обосно
    0.48
    쨔
    0.48
     prest
    0.48
    POSITIVE LOGITS
     trash
    1.92
     garbage
    1.84
     Trash
    1.77
    垃圾
    1.72
     Garbage
    1.68
    trash
    1.67
    Trash
    1.63
     rubbish
    1.61
    garbage
    1.55
     basura
    1.49
    Activations Density 0.144%

    No Known Activations