© 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. 63560
    Prev
    Next
    INDEX
    Explanations

    Universities and Catholic institutions

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

    The neuron simply encodes which exact word (token) it is looking at, effectively acting as a one-hot/vocabulary-index detector rather than responding to any linguistic pattern.

    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
    膾
    -0.80
    Pragma
    -0.77
    愕
    -0.75
     Guernsey
    -0.74
     artysty
    -0.73
    plano
    -0.72
     Partei
    -0.71
     progen
    -0.71
    那種
    -0.71
    }});
    -0.69
    POSITIVE LOGITS
     Jesuit
    1.38
     Catholic
    1.06
    Catholic
    1.03
     като
    1.00
     Loyola
    0.94
     Opus
    0.92
     Catholicism
    0.90
     Villanova
    0.89
     Franciscan
    0.88
    Opus
    0.87
    Activations Density 0.040%

    No Known Activations