© 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. 22-GEMMASCOPE-RES-131K
    7. 26673
    Prev
    Next
    INDEX
    Explanations

    custom security policies

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

    The neuron fires on names of software frameworks, libraries, or configuration elements (e.g. Spring, OAuth, Custom, Security) rather than ordinary prose words.

    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
     each
    -1.34
    illir
    -1.31
     revêtement
    -1.27
     Excelente
    -1.20
    oreo
    -1.19
     Hvor
    -1.19
    raught
    -1.18
    ׇ
    -1.18
     Offisielt
    -1.18
    >';
    -1.16
    POSITIVE LOGITS
     улица
    1.22
     externi
    1.02
    licherweise
    1.02
    懲
    1.01
    ἠ
    1.00
    𝐱
    1.00
    quíni
    0.98
    Nucl
    0.98
     policy
    0.97
    となった
    0.95
    Activations Density 0.002%

    No Known Activations