© 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 EvalsExportsAPI Community BlogPrivacy & TermsContact
    1. Home
    3. Gemma-2-27B
    5. 22-GEMMASCOPE-RES-131K
    7. 38689
    Prev
    Next
    INDEX
    Explanations

    mathematical, code, and scientific terms

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

    the neuron is sensitive to rare or specialized tokens—i.e. low-frequency, often longer or technical words (proper names, class or method identifiers, chemical/biological terms, 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
    χε
    -1.18
     SEMESTER
    -1.15
    lization
    -1.09
     parfüm
    -1.02
    Теперь
    -1.02
    ätigung
    -1.00
    ynomial
    -0.98
    Linki
    -0.96
     maneira
    -0.95
     triko
    -0.95
    POSITIVE LOGITS
     in
    1.11
    !
    0.98
    ,
    0.94
     and
    0.89
    Several
    0.83
    Every
    0.83
     mtu
    0.78
    过期
    0.77
     on
    0.76
    脑海
    0.75
    Activations Density 0.109%

    No Known Activations