© 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. 35794
    Prev
    Next
    INDEX
    Explanations

    best career advice

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

    The neuron detects second‐person nostalgia hooks of the form “If you grew up in … you remember,” i.e. “If you…” style intros aimed at reader recall.

    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
     quickly
    -1.02
     with
    -1.02
     does
    -0.99
     on
    -0.99
     mainly
    -0.99
     swiftly
    -0.99
     settimane
    -0.99
     new
    -0.98
     featuring
    -0.95
     nostro
    -0.94
    POSITIVE LOGITS
     coloriage
    1.07
    explique
    1.06
     normes
    1.05
     δὲ
    1.04
     Estoy
    0.99
     sorgfäl
    0.99
     Excelente
    0.98
     Pued
    0.96
     Ayuda
    0.95
    mujeres
    0.94
    Activations Density 0.051%

    No Known Activations