© 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. Joseph Bloom · Feature Splitting for GPT2-Small
    5. GPT2-Small
    7. Residual Feature Splitting 24576
    9. 8-RES_FS24576-JB
    11. 12723
    Prev
    Next
    INDEX
    Explanations

    names and titles of individuals, such as "Dr. Sarandev Brahambra"

    oai_token-act-pair · gpt-3.5-turboTriggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GPT2-SMALL @ 8-res_fs24576-jb
    Configuration
    jbloom/GPT2-Small-Feature-Splitting-Experiment-Layer-8/blocks.8.hook_resid_pre_24576
    Prompts (Dashboard)
    24,576 prompts, 128 tokens each
    Dataset (Dashboard)
    Skylion007/openwebtext
    Features
    24,576
    Data Type
    torch.float32
    Hook Point
    blocks.8.hook_resid_pre
    Architecture
    standard
    Context Size
    128
    Dataset
    Skylion007/openwebtext
    Hook Point Layer
    8
    Activation Function
    relu
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Neuron Alignment
    Index
    Value
    % of L₁
    481
    +0.12
    0.6%
    409
    +0.11
    0.5%
    447
    +0.11
    0.5%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    232
    +0.12
    0.05
    659
    +0.11
    0.04
    409
    +0.11
    0.04
    Negative Logits
     Codec
    -0.81
     Squirrel
    -0.79
     Caval
    -0.76
     Radar
    -0.75
     Raven
    -0.74
     Rockets
    -0.73
     DD
    -0.73
     Telescope
    -0.72
     Bull
    -0.71
     Pen
    -0.71
    POSITIVE LOGITS
    older
    1.21
    we
    1.14
    so
    1.11
    organ
    1.08
    do
    1.07
    want
    1.06
    altern
    1.05
    to
    1.05
    no
    1.05
    and
    1.04
    Activations Density 0.163%

    No Known Activations