© 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
    2. Gemma-2-27B
    3. 22-GEMMASCOPE-RES-131K
    4. 84190
    Prev
    Next
    INDEX
    Explanations

    strong descriptors

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

    The neuron is looking for hyperbolic intensifiers and glowing promotional language (e.g. “simply,” “amaze,” “like a … dream,” “beautiful”).

    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
     process
    -1.29
     option
    -1.15
    érez
    -1.14
    chapper
    -1.09
     бампер
    -1.08
    杩
    -1.07
     options
    -1.07
     profile
    -1.05
     or
    -1.02
     structure
    -1.02
    POSITIVE LOGITS
    ühen
    1.12
    ׇ
    1.05
     simplemente
    1.04
    #%%
    
    1.04
    issenschaft
    1.02
     solches
    1.02
    たちも
    1.02
    igslist
    1.02
     simply
    1.01
     прекрасно
    1.00
    Activations Density 0.031%

    No Known Activations