© 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. 10-GEMMASCOPE-RES-131K
    7. 75112
    Prev
    Next
    INDEX
    Explanations

    cardinal directions with street names

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

    The neuron detects compass‐point abbreviations (e.g. “E.”, “W.”, “N.”, “East”, “West”) used as directional components in street addresses.

    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
    </strong>
    -2.23
     
    -1.91
    </h3>
    -1.80
    .
    -1.73
    )
    -1.69
     predominant
    -1.39
      
    -1.39
    ↵↵↵
    -1.35
    …”
    -1.31
     since
    -1.30
    POSITIVE LOGITS
    ilares
    1.56
    spedes
    1.55
    alians
    1.54
     chré
    1.53
    gencias
    1.52
     swat
    1.51
    STUDIES
    1.50
    fficients
    1.49
     xxiv
    1.48
    berra
    1.48
    Activations Density 0.023%

    No Known Activations