© 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. 39156
    Prev
    Next
    INDEX
    Explanations

    down or up

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

    The neuron strongly activates on occurrences of the word “down” (often in the phrase “down the …”) indicating a downward direction or movement.

    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
    oléon
    -1.31
     всичко
    -1.29
    ,
    -1.26
    <bos>
    -1.25
     edildi
    -1.18
    ↵↵
    -1.17
    เกี่ยวกับ
    -1.15
     to
    -1.14
    張り
    -1.12
    がなく
    -1.09
    POSITIVE LOGITS
     the
    1.88
     خبرگزاری
    1.42
    を選択します
    1.36
     beeintr
    1.35
     calça
    1.34
     pulseira
    1.34
     various
    1.32
    ternos
    1.31
    必要があります
    1.30
     projeto
    1.30
    Activations Density 0.013%

    No Known Activations