© 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. 10-GEMMASCOPE-RES-131K
    7. 40506
    Prev
    Next
    INDEX
    Explanations

    additive manufacturing, road study

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

    This neuron responds to long sequences of alphanumeric characters in Base64-style blocks, such as those found in cryptographic key or certificate dumps.

    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
    “
    -2.36
    ⢊
    -2.19
     is
    -2.16
     ihre
    -2.08
    涡
    -1.99
    W
    -1.97
     easiest
    -1.88
     of
    -1.85
    J
    -1.80
     THEIR
    -1.80
    POSITIVE LOGITS
    änemark
    3.00
    厛
    2.53
    私には
    2.38
    龘
    2.23
    着一个
    2.19
    言葉を
    2.19
    亹
    2.17
    鵟
    2.11
    瞜
    2.06
    CIÓN
    2.05
    Activations Density 0.017%

    No Known Activations