© 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. 34-GEMMASCOPE-RES-131K
    7. 108978
    Prev
    Next
    INDEX
    Explanations

    movement backward or cost

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

    The neuron strongly activates on instances of the particle “back” when it appears as part of a phrasal verb (e.g. “set you back,” “scale back,” “pull back,” “sits back,” “brushed-back”).

    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
    frontend
    -0.78
    ächtig
    -0.75
    Palo
    -0.74
     Thacker
    -0.73
    lysis
    -0.72
     Frontend
    -0.71
     Andersson
    -0.70
    时的
    -0.69
     Basement
    -0.69
     ندارم
    -0.69
    POSITIVE LOGITS
     back
    6.38
    back
    4.00
    Back
    3.78
    BACK
    3.31
     BACK
    3.20
     zurück
    3.13
     Back
    3.11
     назад
    2.95
     terug
    2.67
     tilbake
    2.56
    Activations Density 0.031%

    No Known Activations