© 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
    2. Gemma-2-27B
    3. 34-GEMMASCOPE-RES-131K
    4. 84062
    Prev
    Next
    INDEX
    Explanations

    Rock, Val, Blanc, Q, TED

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

    The neuron activates on numeric tokens (digit sequences), especially multi-digit numbers like years or numeric 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
    のね
    -0.84
    hörer
    -0.78
    Sqft
    -0.76
    閥
    -0.73
    Entidad
    -0.71
    הח
    -0.70
    Accessing
    -0.69
    קו
    -0.69
    roomId
    -0.69
    consin
    -0.69
    POSITIVE LOGITS
     appellation
    0.88
     Einzel
    0.78
     sub
    0.77
    ű
    0.72
    ผ
    0.72
    ステージ
    0.70
    纷
    0.69
    china
    0.69
    Глав
    0.68
    외
    0.68
    Activations Density 0.021%

    No Known Activations