© 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 EvalsExportsinterp-engineNEWAPI Community BlogPrivacy & TermsContact
    1. Home
    3. Gemma-2-27B
    5. 22-GEMMASCOPE-RES-131K
    7. 65680
    Prev
    Next
    INDEX
    Explanations

    JSON configuration keywords

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

    The neuron responds most strongly to JSON‐style object property names (the quoted keys in JSON).

    oai_token-act-pair · o4-miniTriggered by @jyhe0408
    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
    cisive
    -1.37
     WHOLE
    -1.36
     மற்றும்
    -1.34
     brillant
    -1.33
     idyl
    -1.30
    Transk
    -1.27
    loggen
    -1.20
     hommage
    -1.20
     चीज़ों
    -1.19
    
    -1.19
    POSITIVE LOGITS
    -
    1.53
     required
    1.40
     used
    1.24
     being
    1.23
     taken
    1.19
     when
    1.19
     -
    1.19
     
    1.18
     use
    1.18
     for
    1.18
    Activations Density 0.019%

    No Known Activations