© Neuronpedia 2026
    Privacy & TermsBlogGitHubSlackTwitterContact
    Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    Natural Language
    Autoencoders
    NEW
    Assistant AxisNEWCircuit TracerUPDATESteerSAE EvalsExportsAPI Community BlogPrivacy & TermsContact
    1. Home
    2. OpenAI · Weight-Sparse Transformers
    3. CircuitGPT-Python
    4. Residual Stream
    5. 7-RESID
    6. 1964
    Prev
    Next
    INDEX
    Explanations

    .balance or cmath.sqrt (negative activations)

    python-code · gemini-2.5-flash-lite

    .head attribute access

    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
    _height
    -0.12
    09
    -0.11
    _distance
    -0.11
    _w
    -0.11
     route
    -0.09
    _pro
    -0.09
     w
    -0.04
    '
    0.00
    #
    0.00
    &
    0.00
    POSITIVE LOGITS
    div
    0.22
    ld
    0.21
    _name
    0.18
    ood
    0.17
    _j
    0.16
    ad
    0.16
    add
    0.15
    _i
    0.15
    q
    0.14
    _string
    0.14
    Activations Density 100.000%

    No Known Activations