© 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
    3. Joseph Bloom · Feature Splitting for GPT2-Small
    5. GPT2-Small
    7. Residual Feature Splitting 24576
    9. 8-RES_FS24576-JB
    11. 271
    Prev
    Next
    INDEX
    Explanations

    contact information and details provided in a document

    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
    474
    +0.12
    0.6%
    402
    +0.11
    0.5%
    536
    +0.09
    0.4%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    474
    +0.12
    0.03
    402
    +0.11
    0.03
    726
    +0.09
    0.03
    Negative Logits
    soever
    -0.72
    isa
    -0.68
    conn
    -0.67
    gall
    -0.65
     Pwr
    -0.59
     inval
    -0.59
    Born
    -0.59
     therefore
    -0.57
    Marginal
    -0.57
     moreover
    -0.57
    POSITIVE LOGITS
     sake
    0.90
     details
    0.89
     explanation
    0.87
     reasons
    0.82
     specifics
    0.80
    ummies
    0.78
     inspiration
    0.78
     purposes
    0.77
     clarification
    0.77
     info
    0.72
    Activations Density 0.112%

    No Known Activations