© 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 EvalsExportsAPI Community BlogPrivacy & TermsContact
    1. Home
    3. Gemma-2-27B
    5. 34-GEMMASCOPE-RES-131K
    7. 99959
    Prev
    Next
    INDEX
    Explanations

    code structure, dates, or sequences

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

    The neuron strongly activates on tokens that occur as standalone “header‐like” items—i.e. single words or numbers appearing alone at the start of a paragraph or code block.

    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
    штей
    -0.85
     to
    -0.85
    cartel
    -0.82
     собою
    -0.80
     throughout
    -0.79
    جیب
    -0.77
    ộm
    -0.76
     Trotzdem
    -0.75
    +),
    -0.75
    ตอบ
    -0.75
    POSITIVE LOGITS
    ٹ
    1.11
    ποι
    0.88
     Cec
    0.84
     labdar
    0.82
    সূ
    0.81
    lard
    0.81
     vun
    0.80
     rédig
    0.79
    igrette
    0.79
    HUR
    0.79
    Activations Density 0.004%

    No Known Activations