© 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. 22-GEMMASCOPE-RES-131K
    7. 130310
    Prev
    Next
    INDEX
    Explanations

    The neuron fires on HTML tag syntax—particularly opening‐tag tokens (like “<”, tag names, attribute names, equals signs, and quote delimiters).

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

    HTML table and attributes

    np_acts-logits-general · gemini-2.5-flash-lite
    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
     not
    -1.21
     with
    -1.13
     after
    -1.12
     but
    -1.12
     although
    -1.09
     when
    -1.09
     from
    -1.08
     this
    -1.05
     how
    -1.05
     to
    -0.96
    POSITIVE LOGITS
    GDPR
    1.13
     vestig
    1.02
    🎤
    1.01
    ровок
    0.98
     senza
    0.94
    Atlético
    0.94
     Pilate
    0.94
    ড়
    0.91
     indirec
    0.91
    CUP
    0.89
    Activations Density 0.004%

    No Known Activations