© Neuronpedia 2026
    Privacy & TermsBlogGitHubSlackTwitterContact
    Neuronpedia logo - a computer chip with a rounded viewfinder border around it
    Neuronpedia
    Jacobian LensNEW
    Natural Language
    Autoencoders
    MAY 26    Assistant AxisJAN 26Circuit TracerAPR 26SteerSAE EvalsExportsinterp-engineNEW Community BlogPrivacy & TermsContact
    1. Home
    3. Gemma-3-12B
    5. 31-GEMMASCOPE-2-RES-16K
    7. 4992
    Prev
    Next
    INDEX
    Explanations

    d

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

    This neuron strongly fires on uppercase single‐letter tokens—e.g. “A,” “B,” “C,” “D”—as used to label multiple‐choice answers.

    oai_token-act-pair · o4-mini
    Triggered 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
    Negative Logits
     ReLU
    0.65
     affirmatively
    0.63
     Lemma
    0.61
    ≔
    0.58
     qubit
    0.57
     AcOH
    0.57
     NaN
    0.56
     Holm
    0.55
     Cedex
    0.55
     voxel
    0.55
    POSITIVE LOGITS
    PLEASE
    1.02
     急
    0.96
    !!!!!!!!
    0.89
    NEED
    0.88
    !!!!!!
    0.88
     PLEASE
    0.87
    !!!!!!!!!!!!!!!!
    0.87
    !!!!!
    0.86
    急
    0.86
    !!!!
    0.83
    Activations Density 0.019%

    No Known Activations