© 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. 94015
    Prev
    Next
    INDEX
    Explanations

    github links and attribute assignments

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

    This neuron fires on code‐like markup tokens—specifically XML/HTML tag and attribute names (e.g. “name”, “xmlns”, “Type”) in resource or markup files.

    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
    ursos
    -0.82
    ůr
    -0.81
    gameObject
    -0.79
    onymus
    -0.77
    qk
    -0.76
     Bericht
    -0.76
     consciência
    -0.76
     there
    -0.75
    anias
    -0.74
    uC
    -0.74
    POSITIVE LOGITS
    addError
    1.07
    ]=
    1.03
    ád
    0.95
    =
    0.93
    }=
    0.93
    )=
    0.92
     pierdas
    0.92
    })=
    0.92
     Algérie
    0.91
    PAL
    0.86
    Activations Density 0.001%

    No Known Activations