© 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. 34-GEMMASCOPE-RES-131K
    7. 50315
    Prev
    Next
    INDEX
    Explanations

    money back

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

    The neuron fires strongly on occurrences of “back” (as in money‐back, bought‐back, jobs‐back, refund, etc.), i.e. contexts involving return or restoration.

    oai_token-act-pair · o4-miniTriggered by @jyhe0408
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    google/gemma-scope-27b-pt-res/layer_34/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
    glut
    -0.73
     ono
    -0.69
     København
    -0.67
    Residual
    -0.66
    σει
    -0.66
    得点
    -0.65
    нили
    -0.65
     bos
    -0.65
     Musk
    -0.64
     Samen
    -0.64
    POSITIVE LOGITS
     back
    1.64
     indietro
    1.27
     назад
    1.23
    คืน
    1.23
    回来
    1.17
    back
    1.10
     zurück
    1.09
     terug
    1.05
    回来了
    1.05
     BACK
    1.04
    Activations Density 0.027%

    No Known Activations