© 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. 25878
    Prev
    Next
    INDEX
    Explanations

    unexpected beginnings

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

    The neuron fires strongly on hyperbolic or superlative descriptive words and intensifiers—i.e. exaggerated adjectives/adverbs used for promotional or emphatic effect.

    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
    原始内容存档于
    -1.03
     ktorí
    -1.02
    nejšie
    -1.02
     in
    -1.01
    providedIn
    -1.00
    ce
    -0.92
    ær
    -0.91
    argento
    -0.91
     often
    -0.91
    bl
    -0.90
    POSITIVE LOGITS
     unexpected
    1.08
     unprecedented
    1.08
     colch
    1.06
     giant
    1.05
     possi
    1.05
     ungewöhn
    1.00
     gigantic
    1.00
     formules
    0.99
    unexpected
    0.99
     vinyle
    0.99
    Activations Density 0.091%

    No Known Activations