© 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 EvalsExportsAPI Community BlogPrivacy & TermsContact
    1. Home
    3. Gemma-2-27B
    5. 22-GEMMASCOPE-RES-131K
    7. 27464
    Prev
    Next
    INDEX
    Explanations

    descriptive adjectives

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

    The neuron activates on descriptive scalar adjectives (e.g. sunny, strong, high, cold) that qualify a measurable property.

    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
    farer
    -1.41
     professionelle
    -1.26
     komplette
    -1.20
    farers
    -1.16
    ↵
    -1.16
    Até
    -1.15
     ari
    -1.11
    Õ
    -1.10
    Ótimo
    -1.10
     perfekte
    -1.09
    POSITIVE LOGITS
     and
    1.95
    ness
    1.77
     but
    1.48
     people
    1.47
    過程中
    1.34
     veranderen
    1.34
     fifteen
    1.34
     chcete
    1.34
     yet
    1.33
     sicura
    1.33
    Activations Density 0.085%

    No Known Activations