© 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 EvalsExportsAPI Community BlogPrivacy & TermsContact
    1. Home
    3. Gemma-2-27B
    5. 34-GEMMASCOPE-RES-131K
    7. 1283
    Prev
    Next
    INDEX
    Explanations

    goal orientations

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

    The neuron is picking out content‐bearing (non–stopword) tokens—especially technical, domain‐specific, or otherwise “important” keywords—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
    メソッド
    -0.76
     academic
    -0.70
     Karate
    -0.70
     Verfahren
    -0.69
    howto
    -0.69
     Marquez
    -0.68
    zeman
    -0.67
    ጎ
    -0.67
     academia
    -0.67
     poissons
    -0.66
    POSITIVE LOGITS
     discussione
    0.96
     interviews
    0.95
     Freuden
    0.87
     genesis
    0.79
     Beliefs
    0.79
     MCK
    0.77
    belief
    0.77
     Cobb
    0.76
     beliefs
    0.76
    NCT
    0.76
    Activations Density 0.022%

    No Known Activations