© Neuronpedia 2026
    Privacy & TermsBlogGitHubSlackTwitterContact
    Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    Natural Language
    Autoencoders
    NEW    Assistant AxisNEWCircuit TracerUPDATESteerSAE EvalsExportsAPI Community BlogPrivacy & TermsContact
    1. Home
    3. Gemma-2-27B
    5. 22-GEMMASCOPE-RES-131K
    7. 84227
    Prev
    Next
    INDEX
    Explanations

    forming specific words from adjacent tokens

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

    The neuron flags archaic, first-person prophetic Bible verses—especially lines beginning with “And I saw” or “And I beheld” (e.g. passages from Revelation).

    oai_token-act-pair · o4-miniTriggered by @jyhe0408
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    google/gemma-scope-27b-pt-res/layer_22/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
     mutiara
    -1.27
     gracios
    -1.27
     abbildung
    -1.25
     schöne
    -1.24
     эсте
    -1.23
    ご注意
    -1.20
     jepang
    -1.20
     ayudarte
    -1.20
     Hidup
    -1.20
    Kunden
    -1.19
    POSITIVE LOGITS
    ṝ
    1.27
     démocratique
    1.13
    ,
    1.12
    วะ
    1.08
     démocr
    1.08
     of
    1.03
     饼
    1.00
    ってない
    0.99
    ۜ
    0.99
     वाला
    0.98
    Activations Density 0.007%

    No Known Activations