© 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. 26524
    Prev
    Next
    INDEX
    Explanations

    code constructs and identifiers

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

    The neuron flags tokens that look like programming identifiers or code-specific keywords—e.g. variable/property names, file‐extension literals, XML/HTML tag names—rather than ordinary text.

    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
     cursory
    -1.03
    覗
    -1.02
    特别是
    -1.00
     fundament
    -0.99
    样的
    -0.96
     interminable
    -0.96
     debacle
    -0.95
    并且
    -0.93
     unceasing
    -0.93
     dainty
    -0.90
    POSITIVE LOGITS
    hdysval
    0.99
    TRIES
    0.99
    vable
    0.98
     todavía
    0.97
    März
    0.97
    entukan
    0.96
    Before
    0.94
    '>"
    0.94
     businesses
    0.92
     acheté
    0.92
    Activations Density 0.016%

    No Known Activations