© 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. 127204
    Prev
    Next
    INDEX
    Explanations

    file paths

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

    This neuron detects technical directory or resource folder names in file paths and URLs (e.g., “images,” “scripts,” “media”).

    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
     because
    -2.16
     also
    -2.16
     needed
    -2.08
     already
    -1.84
     didn
    -1.78
     always
    -1.71
     required
    -1.70
     considered
    -1.68
     still
    -1.63
     undoubtedly
    -1.62
    POSITIVE LOGITS
     allerlei
    1.80
    recently
    1.66
     praktisch
    1.55
    anywhere
    1.52
     anún
    1.52
     faciliter
    1.52
     Donc
    1.51
    满脸
    1.50
     óra
    1.49
     véhic
    1.48
    Activations Density 0.032%

    No Known Activations