© 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. 5208
    Prev
    Next
    INDEX
    Explanations

    variable or code punctuation

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

    This neuron spikes on infrequent or unexpected tokens—i.e. the rarer, less common words or punctuation marks that stand out in the text.

    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
    cstring
    0.64
    getattr
    0.61
    Xd
    0.59
     getattr
    0.59
    把
    0.58
    setParameter
    0.57
    beginTransaction
    0.57
    Compute
    0.56
    Expect
    0.56
    行
    0.55
    POSITIVE LOGITS
     राजप
    0.63
     detta
    0.62
    valentine
    0.61
     Selain
    0.60
     এছাড়াও
    0.58
     Snowdon
    0.58
     york
    0.57
     관광
    0.56
     verrez
    0.56
     इससे
    0.55
    Activations Density 0.001%

    No Known Activations