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

    Neuronpedia

    Jacobian LensNEW
    Natural Language
    Autoencoders
    NEW    Assistant AxisNEWCircuit TracerUPDATESteerSAE EvalsExportsinterp-engineNEWAPI Community BlogPrivacy & TermsContact
    1. Home
    3. Gemma-2-27B
    5. 34-GEMMASCOPE-RES-131K
    7. 57757
    Prev
    Next
    INDEX
    Explanations

    Bilas, Bishop, bisphenol

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

    chemical compound names and chemical notation, particularly focusing on prefixes like 'bis', suffixes like '-yne', and chemical formulas.

    oai_token-act-pair · claude-3-7-sonnet-20250219Triggered by @neilrathi

    The neuron is specialized to detect tokens containing digits—e.g. years, numeric values, and other number-based identifiers.

    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
     Versch
    -0.74
    秩序
    -0.70
     around
    -0.70
     Limb
    -0.68
     offenen
    -0.68
    ضاء
    -0.68
     criticise
    -0.67
     given
    -0.66
     Suche
    -0.65
    隊
    -0.65
    POSITIVE LOGITS
     absolument
    0.95
    Bila
    0.95
    важа
    0.86
     MATTERS
    0.84
    taine
    0.81
    盘点
    0.78
    lorette
    0.78
    feels
    0.78
    impro
    0.76
     remarquable
    0.76
    Activations Density 0.031%

    No Known Activations