© 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. OpenAI · Weight-Sparse Transformers
    5. CircuitGPT-Python
    7. Residual Stream
    9. 7-RESID
    11. 1109
    Prev
    Next
    INDEX
    Explanations

    variable `b` followed by `.` (negative activations)

    python-code · gemini-2.5-flash-lite

    accessing class attributes or list elements

    python-code · gemini-2.5-flash-lite
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    Prompts (Dashboard)
    16,384 prompts, 256 tokens each
    Dataset (Dashboard)
    neuronpedia/python-code-simplified
    No Configuration Found
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
    ax
    -0.17
    _max
    -0.11
    (max
    -0.09
    max
    -0.09
    44
    -0.09
     max
    -0.08
     char
    -0.08
     <=
    -0.07
     direction
    -0.05
     None
    -0.03
    POSITIVE LOGITS
    _start
    0.23
     start
    0.22
    _left
    0.20
     left
    0.20
     source
    0.20
    start
    0.20
    .start
    0.20
    irst
    0.19
    left
    0.18
     low
    0.18
    Activations Density 100.000%

    No Known Activations