© 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. 130912
    Prev
    Next
    INDEX
    Explanations

    The neuron fires strongly on semantically rich, content‐bearing tokens—proper nouns, technical terms, and rare topic words.

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

    boxing, driving, assists, latest generation

    np_acts-logits-general · gemini-2.5-flash-lite
    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
    FDRE
    -1.08
    賊
    -1.05
    erape
    -1.05
    rénées
    -1.04
    ikirkan
    -1.02
    思议
    -0.98
    гән
    -0.98
    magyar
    -0.97
    calaure
    -0.96
    łom
    -0.96
    POSITIVE LOGITS
     feveral
    1.10
     lefs
    1.09
    ichtliche
    1.09
     juft
    1.07
     člověk
    1.06
    摂取
    1.05
     जबकि
    1.02
    parently
    1.00
     alha
    1.00
     »>
    0.99
    Activations Density 0.276%

    No Known Activations