© 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. 5071
    Prev
    Next
    INDEX
    Explanations

    place to stay for a day

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

    The neuron detects marketing‐style call‐to‐action language around travel and accommodation (e.g. verbs like “come,” “stay,” “book,” “travelling”).

    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
     가서
    0.60
    报名
    0.55
     рано
    0.55
    午後
    0.54
     രംഗ
    0.51
    日上午
    0.51
    ង
    0.51
    摸
    0.50
    下午
    0.50
    แสง
    0.49
    POSITIVE LOGITS
     visiting
    1.45
     overnight
    1.29
     Visiting
    1.26
     visits
    1.23
     transient
    1.21
     stays
    1.20
    Visiting
    1.20
     stay
    1.19
     transitory
    1.18
     visit
    1.17
    Activations Density 0.421%

    No Known Activations