© 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. 64859
    Prev
    Next
    INDEX
    Explanations

    new boardwalk, q:, ut-, biofeedback

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

    The neuron strongly activates on sequences of hexadecimal digits (0–9, A–F) and thus is detecting hex constants or hex 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_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
    strukt
    -1.16
    They
    -1.08
    Initially
    -1.02
    therape
    -1.01
     jeudi
    -1.00
     awal
    -0.99
    Apart
    -0.98
    Both
    -0.98
    Aside
    -0.98
    kundige
    -0.97
    POSITIVE LOGITS
     then
    1.33
     there
    1.23
     imediatamente
    1.13
    愜
    1.09
     Scotsman
    1.09
     belangrijke
    1.05
     Boud
    1.05
     Flock
    1.03
     protecting
    1.00
     aceptación
    1.00
    Activations Density 0.003%

    No Known Activations