© 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. 93354
    Prev
    Next
    INDEX
    Explanations

    Australian cultural and agricultural issues

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

    The neuron selectively lights up on semantically rich content words—especially domain-specific nouns, verbs, adjectives, and numerals—while ignoring common function words.

    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
    颇
    -0.94
    Correlation
    -0.91
    berichte
    -0.90
     VIC
    -0.89
     ผู้
    -0.86
     cacao
    -0.86
     диван
    -0.85
    МИ
    -0.83
    收到
    -0.82
     Culti
    -0.82
    POSITIVE LOGITS
     enhancing
    0.88
     diminish
    0.87
     genuinely
    0.85
     Gemein
    0.85
     enhanced
    0.84
    ---|
    0.84
    であり
    0.83
     our
    0.83
     both
    0.81
    bieron
    0.81
    Activations Density 0.054%

    No Known Activations