© 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
    3. Gemma-2-27B
    5. 22-GEMMASCOPE-RES-131K
    7. 79662
    Prev
    Next
    INDEX
    Explanations

    writing, thing, is, got, things, :, ks, xml

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

    The neuron fires strongly on tokens typically used in enthusiastic positive feedback—expressions of thanks, praise, and approval in a comment or post.

    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
     three
    -1.46
     two
    -1.31
     four
    -1.29
     without
    -1.27
    única
    -1.24
     Anggota
    -1.15
     five
    -1.14
     δύο
    -1.13
     čase
    -1.13
    pasti
    -1.12
    POSITIVE LOGITS
     tinham
    1.18
     this
    1.17
    ↵
    1.14
     precisam
    1.13
     terrific
    1.07
     here
    1.01
    鲼
    1.01
     encontram
    0.99
    
    0.99
    this
    0.97
    Activations Density 0.021%

    No Known Activations