© 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. 34-GEMMASCOPE-RES-131K
    7. 77878
    Prev
    Next
    INDEX
    Explanations

    health-related activities and issues

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

    The neuron detects the word "related" when it appears in hyphenated terms or phrases, especially in academic or technical contexts.

    oai_token-act-pair · claude-3-7-sonnet-20250219Triggered by @neilrathi

    The neuron specifically fires on the word “related,” especially when it appears as part of a hyphenated modifier (e.g. “X-related”).

    oai_token-act-pair · o4-miniTriggered by @jyhe0408
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    google/gemma-scope-27b-pt-res/layer_34/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
    ону
    -0.97
     sorgen
    -0.96
     kult
    -0.96
     supplementing
    -0.94
     modalités
    -0.94
    znál
    -0.94
    inawa
    -0.93
    क्या
    -0.92
    islamic
    -0.91
    狳
    -0.91
    POSITIVE LOGITS
     activities
    1.98
     issues
    1.47
     tasks
    1.39
     items
    1.38
     topics
    1.37
     events
    1.28
     services
    1.25
     stuff
    1.22
     projects
    1.22
     matters
    1.20
    Activations Density 0.046%

    No Known Activations