© 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. 10-GEMMASCOPE-RES-131K
    7. 43114
    Prev
    Next
    INDEX
    Explanations

    paths

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

    The neuron fires on occurrences of “paths” in code, docs, or config—i.e. references to filesystem or module path variables.

    oai_token-act-pair · o4-miniTriggered by @jyhe0408
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    google/gemma-scope-27b-pt-res/layer_10/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
    er
    -3.22
    u
    -3.19
    y
    -2.84
    erintah
    -2.81
    igrette
    -2.80
    ⎆
    -2.77
     ainfi
    -2.73
    وضح
    -2.72
    ardige
    -2.69
    ed
    -2.61
    POSITIVE LOGITS
    "
    3.25
    Initially
    3.02
    Previously
    2.88
    Each
    2.81
    5
    2.67
    AN
    2.66
    很不
    2.64
    Instead
    2.64
    Additionally
    2.58
    Therefore
    2.53
    Activations Density 0.007%

    No Known Activations