© 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. 68473
    Prev
    Next
    INDEX
    Explanations

    X or Y

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

    The neuron fires on words that express spatial location or direction (e.g. “on,” “side,” “town,” “other,” “live,” “close”), basically highlighting prepositional phrases that describe where something is.

    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
     inoxydable
    -0.97
     tellement
    -0.92
     Ș
    -0.92
    Cryptography
    -0.92
    ał
    -0.91
    aroa
    -0.90
     présidenti
    -0.88
    ieteur
    -0.87
     PSOE
    -0.86
     descrip
    -0.85
    POSITIVE LOGITS
     or
    1.79
     any
    1.12
     as
    0.98
     when
    0.92
     будь
    0.86
     just
    0.85
     here
    0.85
     Compared
    0.84
     hundreds
    0.83
     this
    0.81
    Activations Density 0.027%

    No Known Activations