© 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. 84315
    Prev
    Next
    INDEX
    Explanations

    positive sentiment expression

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

    The neuron detects strong positive appraisal verbs used to endorse or recommend something (e.g. “love,” “enjoy,” “appreciate”).

    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
    それほど
    -1.05
    imakasih
    -1.02
    usions
    -1.01
    )}{(
    -1.00
    してしまった
    -1.00
     dergleichen
    -0.98
    ksha
    -0.95
     zodat
    -0.95
     Sometimes
    -0.93
     There
    -0.93
    POSITIVE LOGITS
     if
    1.67
     progrès
    1.23
    もし
    1.12
     these
    1.11
    白菜
    1.09
    inį
    1.05
    strö
    1.02
     your
    1.02
     menyenangkan
    1.01
    この
    1.00
    Activations Density 0.022%

    No Known Activations