© 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. 22-GEMMASCOPE-RES-131K
    7. 128682
    Prev
    Next
    INDEX
    Explanations

    The neuron is picking up non‐word tokens—things like punctuation marks, digits, symbols, and other non-alphabetic characters.

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

    that

    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
    asam
    -1.10
    :"-
    -1.08
     komfor
    -0.98
    FDRE
    -0.88
     balans
    -0.88
    pannt
    -0.87
    ciech
    -0.85
    szönöm
    -0.83
    hopal
    -0.83
    所以
    -0.82
    POSITIVE LOGITS
    那是
    1.28
     that
    1.23
    実は
    1.10
     That
    1.05
    That
    1.00
     THAT
    1.00
     Thats
    0.98
     thats
    0.98
     those
    0.98
    はその
    0.98
    Activations Density 0.079%

    No Known Activations