© 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. 423
    Prev
    Next
    INDEX
    Explanations

    names of universities and sports teams

    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
    161
    +0.10
    0.5%
    481
    +0.10
    0.4%
    377
    +0.10
    0.4%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    575
    +0.10
    0.03
    132
    +0.10
    0.03
    401
    +0.10
    0.03
    Negative Logits
    Effective
    -0.96
    ij士
    -0.84
    £ı
    -0.83
    é¾
    -0.79
    -+-+
    -0.79
    ãĥ¼ãĥĨ
    -0.77
     Effective
    -0.76
    ãĥ¯ãĥ³
    -0.76
    ERO
    -0.75
    OLOG
    -0.75
    POSITIVE LOGITS
    hood
    1.36
    maid
    1.11
    folk
    1.06
     who
    0.89
     whom
    0.88
    child
    0.84
     born
    0.82
    friend
    0.81
     loves
    0.81
     rejoice
    0.81
    Activations Density 0.140%

    No Known Activations