© 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. 59863
    Prev
    Next
    INDEX
    Explanations

    multi and single variants

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

    This neuron detects occurrences of the “multi-” prefix (e.g. in compound words like “multi-copy,” “multi-member,” “multiple,” etc.).

    oai_token-act-pair · o4-miniTriggered by @jyhe0408

    terms indicating multiplicity or numerical count, especially hyphenated constructions denoting numbers of units or copies.

    oai_token-act-pair · gpt-5Triggered by @jyhe0408

    the prefix "multi-" in compound words and phrases.

    oai_token-act-pair · claude-4-5-sonnetTriggered 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
    有意思
    -0.79
     mettre
    -0.76
     Α
    -0.74
     病
    -0.68
    鐲
    -0.68
     donne
    -0.66
     Celebrating
    -0.66
    ਰ
    -0.66
     expresó
    -0.66
    engers
    -0.66
    POSITIVE LOGITS
     peaked
    0.84
     valued
    0.83
    sbereich
    0.82
     occupancy
    0.78
    лар
    0.78
     mandibular
    0.78
     or
    0.76
    谢
    0.73
    InRange
    0.73
    Multi
    0.73
    Activations Density 0.085%

    No Known Activations