© 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
    2. Gemma-2-27B
    3. 10-GEMMASCOPE-RES-131K
    4. 8751
    Prev
    Next
    INDEX
    Explanations

    KwaZulu-Natal, Dalaca, Kabeiro, Kailash, Berlin

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

    The neuron fires on the initial subtoken of capitalized named entities—e.g. the first piece of place-names, species names, acronyms, or other proper nouns.

    oai_token-act-pair · o4-miniTriggered by @jyhe0408

    segments of proper nouns and acronyms, especially unusual or non-English capitalized syllables.

    oai_token-act-pair · gpt-5Triggered by @jyhe0408

    proper nouns or names, particularly those from non-English languages or geographic locations.

    oai_token-act-pair · claude-4-5-sonnetTriggered 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
    alakip
    -2.11
     of
    -1.90
    ſſen
    -1.84
    ikuti
    -1.82
    becue
    -1.67
    𝒕
    -1.67
    _________
    -1.66
    ___________
    -1.65
     fantastisk
    -1.65
    
    -1.64
    POSITIVE LOGITS
     /
    1.98
     (
    1.98
     it
    1.71
     capa
    1.64
    ]--;
    1.59
    xically
    1.56
     .....
    1.50
     gets
    1.50
       
    1.48
     števil
    1.46
    Activations Density 0.002%

    No Known Activations