© 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. 10-GEMMASCOPE-RES-131K
    7. 111754
    Prev
    Next
    INDEX
    Explanations

    JSON objects with lists or array starts `[`

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

    colons in JSON key-value pair syntax.

    oai_token-act-pair · claude-4-5-sonnetTriggered by @bfgcn7frf5

    The neuron selectively fires on the “:” tokens that serve as key‐value separators in JSON structures.

    oai_token-act-pair · o4-miniTriggered by @jyhe0408
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    google/gemma-scope-27b-pt-res/layer_10/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
    This
    -2.42
     Though
    -2.34
     That
    -2.33
    </h4>
    -2.28
    These
    -2.22
    3
    -2.19
    Only
    -2.19
     Regardless
    -2.17
    They
    -2.14
     However
    -2.14
    POSITIVE LOGITS
     distritos
    2.25
    silly
    2.05
     masculinos
    2.00
    decide
    1.98
     relativamente
    1.96
     poderos
    1.96
     virksom
    1.96
     respectivas
    1.96
     短发
    1.95
     trzec
    1.93
    Activations Density 0.007%

    No Known Activations