© 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. 10-GEMMASCOPE-RES-131K
    7. 58589
    Prev
    Next
    INDEX
    Explanations

    chemical nomenclature and names

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

    foreign language characters and uncommon letter combinations.

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

    This neuron flags uncommon proper names and specialized terminology—e.g. person names, foreign words, acronyms, and chemical compound names.

    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
    衕
    -2.72
    豨
    -2.70
    îte
    -2.61
     kulinar
    -2.59
    鸸
    -2.56
    お楽しみに
    -2.55
    聼
    -2.55
     ervaren
    -2.55
     说
    -2.52
    
    -2.52
    POSITIVE LOGITS
    d
    3.53
    }
    3.53
    if
    3.44
    ",
    2.80
    us
    2.75
    //
    2.70
    ',
    2.63
    !!!!!!
    2.61
    s
    2.61
    “
    2.55
    Activations Density 0.058%

    No Known Activations