© 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. 116624
    Prev
    Next
    INDEX
    Explanations

    under one or more contributor license agreements

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

    The neuron is highly activated by standalone numeric tokens (digits/slashes) in code comments—e.g. version numbers, years, and other numeric constants.

    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
     申
    -0.75
    𝚟
    -0.73
    umato
    -0.70
     antennae
    -0.69
     outlier
    -0.69
    Tara
    -0.67
    izace
    -0.66
    amounts
    -0.65
     amounts
    -0.64
    Camping
    -0.64
    POSITIVE LOGITS
     sterling
    0.76
     הייתה
    0.69
     Tone
    0.69
     irradiated
    0.67
    Tone
    0.66
    êm
    0.65
    FOUR
    0.65
     bezpiecze
    0.65
    ică
    0.64
     Lunes
    0.64
    Activations Density 0.123%

    No Known Activations