© 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. 34-GEMMASCOPE-RES-131K
    7. 63568
    Prev
    Next
    INDEX
    Explanations

    self in class definitions

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

    This neuron fires on code tokens that define or name model parameters or layers—e.g. layer names in config (“name: 'conv1'”) and Python attribute assignments to self (self.transform, self.root, etc.).

    oai_token-act-pair · o4-miniTriggered by @jyhe0408
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    google/gemma-scope-27b-pt-res/layer_34/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
    ğun
    -0.88
     Kedua
    -0.82
     realizan
    -0.77
     두
    -0.77
     verbunden
    -0.75
     véritable
    -0.75
     nächsten
    -0.75
    Marcin
    -0.74
     vuelven
    -0.74
     vinden
    -0.73
    POSITIVE LOGITS
    register
    0.99
     register
    0.98
    REGISTER
    0.94
     матч
    0.91
     регистрации
    0.86
    )=\
    0.86
     registers
    0.84
    ়
    0.82
     smiles
    0.81
    Register
    0.80
    Activations Density 0.002%

    No Known Activations