© 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. 22-GEMMASCOPE-RES-131K
    7. 69166
    Prev
    Next
    INDEX
    Explanations

    births and findings

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

    The neuron is detecting numeric tokens or numerical values in the text.

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

    line breaks or paragraph breaks in the text, particularly those separating sections or introducing new topics.

    oai_token-act-pair · claude-4-5-sonnetTriggered by @jyhe0408

    indicators of encyclopedia-style structure—section headers and tokens that frame factual, definitional, or biographical statements.

    oai_token-act-pair · gpt-5Triggered by @jyhe0408
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    google/gemma-scope-27b-pt-res/layer_22/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
     another
    -1.12
     くん
    -0.95
    🔑
    -0.94
    تاح
    -0.90
     if
    -0.90
    {(-
    -0.88
    propagation
    -0.87
     kulinar
    -0.87
    pullover
    -0.86
    примеча
    -0.85
    POSITIVE LOGITS
     în
    1.07
     también
    1.05
    чів
    1.02
     also
    1.02
    illi
    0.94
     juga
    0.93
     довольно
    0.92
    也得
    0.91
    ultura
    0.91
    :\
    0.90
    Activations Density 0.001%

    No Known Activations