© Neuronpedia 2026
    Privacy & TermsBlogGitHubSlackTwitterContact
    Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    Natural Language
    Autoencoders
    NEW    Assistant AxisNEWCircuit TracerUPDATESteerSAE EvalsExportsAPI Community BlogPrivacy & TermsContact
    1. Home
    3. Gemma-2-27B
    5. 34-GEMMASCOPE-RES-131K
    7. 83838
    Prev
    Next
    INDEX
    Explanations

    assume

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

    The neuron responds strongly to very high‐frequency tokens—common short words (articles, prepositions, determiners) and frequent code/macros—i.e. typical “stop words” and standard programming directives.

    oai_token-act-pair · o4-miniTriggered by @jyhe0408
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    google/gemma-scope-27b-pt-res/layer_34/width_131k
    Prompts (Dashboard)
    24,576 prompts, 128 tokens each
    Dataset (Dashboard)
    monology/pile-uncopyrighted
    No Configuration Found
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
    లి
    -0.85
    arci
    -0.82
    Jäh
    -0.81
     drumming
    -0.79
    limsy
    -0.79
     раньше
    -0.76
     Բ
    -0.76
    hasMoreElements
    -0.74
    orse
    -0.74
     запо
    -0.74
    POSITIVE LOGITS
    ဏ
    0.85
     Bekasi
    0.81
    relations
    0.80
    hehe
    0.78
     coklat
    0.77
    missing
    0.74
     редко
    0.72
     biri
    0.72
     pues
    0.71
     segü
    0.71
    Activations Density 0.032%

    No Known Activations