© 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. 39695
    Prev
    Next
    INDEX
    Explanations

    technical terms followed by specifics

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

    The neuron strongly activates on superlative or extremal qualifiers—words that specify “best,” “shortest,” “optimal,” etc.

    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
    最近は
    -1.31
     vaksin
    -1.28
     kompeti
    -1.26
     Öz
    -1.16
    <bos>
    -1.15
     bobina
    -1.11
     métiers
    -1.11
    ख
    -1.10
     verás
    -1.07
    जन
    -1.06
    POSITIVE LOGITS
     it
    1.40
     which
    1.17
    août
    1.16
    It
    1.16
     unserer
    1.13
     Information
    1.13
    X
    1.12
    為に
    1.12
     agricultural
    1.11
    N
    1.11
    Activations Density 0.155%

    No Known Activations