© 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. 128822
    Prev
    Next
    INDEX
    Explanations

    The neuron detects hedging or qualification phrases centered on “depends on” (e.g. “It depends on …”).

    oai_token-act-pair · o4-miniTriggered by @jyhe0408

    depends on circumstances

    np_acts-logits-general · gemini-2.5-flash-lite
    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
     mediante
    -1.08
    nages
    -1.07
    inawa
    -1.01
     while
    -1.00
    ixante
    -0.97
     jedynie
    -0.96
    vább
    -0.96
    leqq
    -0.95
     أجل
    -0.95
     komt
    -0.94
    POSITIVE LOGITS
     circumstances
    1.88
     situation
    1.37
     what
    1.16
     how
    1.15
    ğından
    1.14
     particular
    1.13
     your
    1.13
    情况
    1.13
     circunstancias
    1.11
    How
    1.09
    Activations Density 0.021%

    No Known Activations