© 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. 10-GEMMASCOPE-RES-131K
    7. 14034
    Prev
    Next
    INDEX
    Explanations

    programming, code, and commands

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

    The neuron fires strongly on uncommon or "out‐of‐the-ordinary" tokens—e.g. function names, mathematical operators, and other rare or non-standard symbols/characters.

    oai_token-act-pair · o4-miniTriggered by @jyhe0408
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    google/gemma-scope-27b-pt-res/layer_10/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
    躂
    -2.38
    .”
    -2.33
     たく
    -2.30
     GOtt
    -2.28
     すぐ
    -2.27
    >
    -2.20
    噤
    -2.17
     evidenced
    -2.14
    ᨗ
    -2.13
     哪些
    -2.11
    POSITIVE LOGITS
    s
    3.39
    了一个
    3.09
     When
    3.09
    </h3>
    2.83
     машине
    2.72
    ༈
    2.67
    噐
    2.55
    嫆
    2.52
    蝋
    2.50
     начинает
    2.48
    Activations Density 0.008%

    No Known Activations